{ "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": 4.835177421569824, "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": 4.588673114776611, "learning_rate": 4.2390843577787196e-10, "loss": 0.8329, "mean_token_accuracy": 0.765798807144165, "num_tokens": 80419.0, "step": 2 }, { "epoch": 0.0003816308357715303, "ewc_loss": 4.9682480351975755e-15, "ewc_loss_diag": 6.132518538121134e-19, "ewc_loss_parallel": 4.9060148304969076e-18, "grad_norm": 4.725718975067139, "learning_rate": 8.478168715557439e-10, "loss": 0.7225, "mean_token_accuracy": 0.7960407137870789, "num_tokens": 118717.0, "step": 3 }, { "epoch": 0.0005088411143620405, "ewc_loss": 2.0694557179012918e-13, "ewc_loss_diag": 5.0523821237824507e-17, "ewc_loss_parallel": 2.0122792321330962e-16, "grad_norm": 5.366392135620117, "learning_rate": 1.271725307333616e-09, "loss": 0.814, "mean_token_accuracy": 0.7712171077728271, "num_tokens": 150155.0, "step": 4 }, { "epoch": 0.0006360513929525506, "ewc_loss": 1.6910917111090384e-12, "ewc_loss_diag": 1.6930901125533637e-15, "ewc_loss_parallel": 1.519617764955683e-15, "grad_norm": 4.306911468505859, "learning_rate": 1.6956337431114878e-09, "loss": 0.7917, "mean_token_accuracy": 0.7746701240539551, "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": 5.211915969848633, "learning_rate": 2.1195421788893596e-09, "loss": 0.7895, "mean_token_accuracy": 0.7781615853309631, "num_tokens": 227640.0, "step": 6 }, { "epoch": 0.0008904719501335708, "ewc_loss": 1.7053025658242404e-11, "ewc_loss_diag": 1.1213252548714081e-14, "ewc_loss_parallel": 1.587618925213974e-14, "grad_norm": 4.833097457885742, "learning_rate": 2.543450614667232e-09, "loss": 0.816, "mean_token_accuracy": 0.7748944759368896, "num_tokens": 265114.0, "step": 7 }, { "epoch": 0.001017682228724081, "ewc_loss": 7.685230229981244e-11, "ewc_loss_diag": 3.397282455352979e-14, "ewc_loss_parallel": 7.37188088351104e-14, "grad_norm": 4.937308311462402, "learning_rate": 2.967359050445104e-09, "loss": 0.7584, "mean_token_accuracy": 0.7886372804641724, "num_tokens": 299865.0, "step": 8 }, { "epoch": 0.001144892507314591, "ewc_loss": 1.191438059322536e-10, "ewc_loss_diag": 5.1958437552457326e-14, "ewc_loss_parallel": 1.1368683772161603e-13, "grad_norm": 4.57670783996582, "learning_rate": 3.3912674862229757e-09, "loss": 0.814, "mean_token_accuracy": 0.7733403444290161, "num_tokens": 342063.0, "step": 9 }, { "epoch": 0.0012721027859051012, "ewc_loss": 2.2919266484677792e-10, "ewc_loss_diag": 1.2878587085651816e-13, "ewc_loss_parallel": 2.1671553440683056e-13, "grad_norm": 5.45999002456665, "learning_rate": 3.815175922000847e-09, "loss": 0.8641, "mean_token_accuracy": 0.7634329199790955, "num_tokens": 374864.0, "step": 10 }, { "epoch": 0.0013993130644956112, "ewc_loss": 6.511982064694166e-10, "ewc_loss_diag": 2.2115642650533118e-13, "ewc_loss_parallel": 6.288303211476887e-13, "grad_norm": 4.309223651885986, "learning_rate": 4.239084357778719e-09, "loss": 0.7764, "mean_token_accuracy": 0.7755395174026489, "num_tokens": 416605.0, "step": 11 }, { "epoch": 0.0015265233430861213, "ewc_loss": 8.803908713161945e-10, "ewc_loss_diag": 3.126388037344441e-13, "ewc_loss_parallel": 8.490985692333197e-13, "grad_norm": 5.464030742645264, "learning_rate": 4.662992793556591e-09, "loss": 0.8347, "mean_token_accuracy": 0.7690819501876831, "num_tokens": 448798.0, "step": 12 }, { "epoch": 0.0016537336216766315, "ewc_loss": 1.1132215149700642e-09, "ewc_loss_diag": 3.872457909892546e-13, "ewc_loss_parallel": 1.0800249583553523e-12, "grad_norm": 5.444677352905273, "learning_rate": 5.086901229334464e-09, "loss": 0.8738, "mean_token_accuracy": 0.7543173432350159, "num_tokens": 480084.0, "step": 13 }, { "epoch": 0.0017809439002671415, "ewc_loss": 3.434251993894577e-09, "ewc_loss_diag": 1.0729195309977513e-12, "ewc_loss_parallel": 3.339550858072471e-12, "grad_norm": 4.387409210205078, "learning_rate": 5.510809665112336e-09, "loss": 0.8055, "mean_token_accuracy": 0.7797832489013672, "num_tokens": 524543.0, "step": 14 }, { "epoch": 0.0019081541788576518, "ewc_loss": 5.122274160385132e-09, "ewc_loss_diag": 1.5987211554602254e-12, "ewc_loss_parallel": 4.973799150320701e-12, "grad_norm": 4.792224407196045, "learning_rate": 5.934718100890208e-09, "loss": 0.7519, "mean_token_accuracy": 0.7882855534553528, "num_tokens": 563314.0, "step": 15 }, { "epoch": 0.002035364457448162, "ewc_loss": 6.199115887284279e-09, "ewc_loss_diag": 1.7763568394002505e-12, "ewc_loss_parallel": 6.0254023992456496e-12, "grad_norm": 5.068251609802246, "learning_rate": 6.3586265366680796e-09, "loss": 0.8422, "mean_token_accuracy": 0.7626417875289917, "num_tokens": 598421.0, "step": 16 }, { "epoch": 0.002162574736038672, "ewc_loss": 7.916241884231567e-09, "ewc_loss_diag": 2.2168933355715126e-12, "ewc_loss_parallel": 7.673861546209082e-12, "grad_norm": 5.057251453399658, "learning_rate": 6.782534972445951e-09, "loss": 0.832, "mean_token_accuracy": 0.7628871202468872, "num_tokens": 634690.0, "step": 17 }, { "epoch": 0.002289785014629182, "ewc_loss": 8.789356797933578e-09, "ewc_loss_diag": 2.6858515411731787e-12, "ewc_loss_parallel": 8.526512829121202e-12, "grad_norm": 5.119034290313721, "learning_rate": 7.206443408223823e-09, "loss": 0.869, "mean_token_accuracy": 0.7597325444221497, "num_tokens": 674653.0, "step": 18 }, { "epoch": 0.0024169952932196924, "ewc_loss": 2.3515895009040833e-08, "ewc_loss_diag": 9.833911462919787e-12, "ewc_loss_parallel": 2.262368070660159e-11, "grad_norm": 5.048445701599121, "learning_rate": 7.630351844001695e-09, "loss": 0.8008, "mean_token_accuracy": 0.777197003364563, "num_tokens": 708238.0, "step": 19 }, { "epoch": 0.0025442055718102024, "ewc_loss": 3.5157427191734314e-08, "ewc_loss_diag": 1.2107648217352107e-11, "ewc_loss_parallel": 3.410605131648481e-11, "grad_norm": 4.629537105560303, "learning_rate": 8.054260279779567e-09, "loss": 0.7823, "mean_token_accuracy": 0.778534471988678, "num_tokens": 749312.0, "step": 20 }, { "epoch": 0.0026714158504007124, "ewc_loss": 4.1676685214042664e-08, "ewc_loss_diag": 1.3926637620897964e-11, "ewc_loss_parallel": 4.0245140553452075e-11, "grad_norm": 5.426647186279297, "learning_rate": 8.478168715557438e-09, "loss": 0.8525, "mean_token_accuracy": 0.7663077116012573, "num_tokens": 783532.0, "step": 21 }, { "epoch": 0.0027986261289912225, "ewc_loss": 4.6566128730773926e-08, "ewc_loss_diag": 1.48929757415317e-11, "ewc_loss_parallel": 4.524736141320318e-11, "grad_norm": 5.1087822914123535, "learning_rate": 8.902077151335311e-09, "loss": 0.8304, "mean_token_accuracy": 0.7731561064720154, "num_tokens": 817429.0, "step": 22 }, { "epoch": 0.0029258364075817325, "ewc_loss": 5.448237061500549e-08, "ewc_loss_diag": 1.7621459846850485e-11, "ewc_loss_parallel": 5.2523319027386606e-11, "grad_norm": 4.8307881355285645, "learning_rate": 9.325985587113182e-09, "loss": 0.7634, "mean_token_accuracy": 0.7884426116943359, "num_tokens": 853964.0, "step": 23 }, { "epoch": 0.0030530466861722425, "ewc_loss": 6.007030606269836e-08, "ewc_loss_diag": 1.944044925039634e-11, "ewc_loss_parallel": 5.820766091346741e-11, "grad_norm": 5.45570182800293, "learning_rate": 9.749894022891054e-09, "loss": 0.8278, "mean_token_accuracy": 0.7670427560806274, "num_tokens": 885070.0, "step": 24 }, { "epoch": 0.003180256964762753, "ewc_loss": 6.472691893577576e-08, "ewc_loss_diag": 2.114575181622058e-11, "ewc_loss_parallel": 6.230038707144558e-11, "grad_norm": 4.829650402069092, "learning_rate": 1.0173802458668929e-08, "loss": 0.8797, "mean_token_accuracy": 0.7512497901916504, "num_tokens": 926893.0, "step": 25 }, { "epoch": 0.003307467243353263, "ewc_loss": 1.471489667892456e-07, "ewc_loss_diag": 6.730260793119669e-11, "ewc_loss_parallel": 1.4097167877480388e-10, "grad_norm": 4.887539863586426, "learning_rate": 1.05977108944468e-08, "loss": 0.8161, "mean_token_accuracy": 0.7676177620887756, "num_tokens": 964773.0, "step": 26 }, { "epoch": 0.003434677521943773, "ewc_loss": 2.2444874048233032e-07, "ewc_loss_diag": 8.685674401931465e-11, "ewc_loss_parallel": 2.1645973902195692e-10, "grad_norm": 4.74567174911499, "learning_rate": 1.1021619330224672e-08, "loss": 0.7691, "mean_token_accuracy": 0.784824788570404, "num_tokens": 1002725.0, "step": 27 }, { "epoch": 0.003561887800534283, "ewc_loss": 2.7008354663848877e-07, "ewc_loss_diag": 9.640643838793039e-11, "ewc_loss_parallel": 2.601154847070575e-10, "grad_norm": 4.973435878753662, "learning_rate": 1.1445527766002543e-08, "loss": 0.8732, "mean_token_accuracy": 0.7579166889190674, "num_tokens": 1040296.0, "step": 28 }, { "epoch": 0.003689098079124793, "ewc_loss": 2.980232238769531e-07, "ewc_loss_diag": 1.0550138540565968e-10, "ewc_loss_parallel": 2.874003257602453e-10, "grad_norm": 4.430424690246582, "learning_rate": 1.1869436201780416e-08, "loss": 0.7656, "mean_token_accuracy": 0.7853502035140991, "num_tokens": 1081711.0, "step": 29 }, { "epoch": 0.0038163083577153036, "ewc_loss": 3.166496753692627e-07, "ewc_loss_diag": 1.1141310096718371e-10, "ewc_loss_parallel": 3.055902197957039e-10, "grad_norm": 4.813132286071777, "learning_rate": 1.2293344637558287e-08, "loss": 0.8119, "mean_token_accuracy": 0.7750792503356934, "num_tokens": 1120556.0, "step": 30 }, { "epoch": 0.003943518636305814, "ewc_loss": 3.427267074584961e-07, "ewc_loss_diag": 1.255102688446641e-10, "ewc_loss_parallel": 3.292370820418e-10, "grad_norm": 4.7828874588012695, "learning_rate": 1.2717253073336159e-08, "loss": 0.7639, "mean_token_accuracy": 0.7893338203430176, "num_tokens": 1157723.0, "step": 31 }, { "epoch": 0.004070728914896324, "ewc_loss": 3.6694109439849854e-07, "ewc_loss_diag": 1.382431946694851e-10, "ewc_loss_parallel": 3.54702933691442e-10, "grad_norm": 4.678961753845215, "learning_rate": 1.314116150911403e-08, "loss": 0.8102, "mean_token_accuracy": 0.7737858891487122, "num_tokens": 1197879.0, "step": 32 }, { "epoch": 0.004197939193486834, "ewc_loss": 3.91155481338501e-07, "ewc_loss_diag": 1.4370016288012266e-10, "ewc_loss_parallel": 3.765308065339923e-10, "grad_norm": 4.760921478271484, "learning_rate": 1.3565069944891903e-08, "loss": 0.8107, "mean_token_accuracy": 0.7750930786132812, "num_tokens": 1237342.0, "step": 33 }, { "epoch": 0.004325149472077344, "ewc_loss": 4.246830940246582e-07, "ewc_loss_diag": 1.546140993013978e-10, "ewc_loss_parallel": 4.092726157978177e-10, "grad_norm": 4.4958577156066895, "learning_rate": 1.3988978380669775e-08, "loss": 0.7821, "mean_token_accuracy": 0.7820889949798584, "num_tokens": 1280197.0, "step": 34 }, { "epoch": 0.004452359750667854, "ewc_loss": 5.140900611877441e-07, "ewc_loss_diag": 1.9463186617940664e-10, "ewc_loss_parallel": 4.94765117764473e-10, "grad_norm": 4.62913703918457, "learning_rate": 1.4412886816447646e-08, "loss": 0.7641, "mean_token_accuracy": 0.7850255370140076, "num_tokens": 1318625.0, "step": 35 }, { "epoch": 0.004579570029258364, "ewc_loss": 1.087784767150879e-06, "ewc_loss_diag": 5.093170329928398e-10, "ewc_loss_parallel": 1.0331859812140465e-09, "grad_norm": 4.822289943695068, "learning_rate": 1.4836795252225519e-08, "loss": 0.8219, "mean_token_accuracy": 0.7724101543426514, "num_tokens": 1356868.0, "step": 36 }, { "epoch": 0.004706780307848874, "ewc_loss": 1.5050172805786133e-06, "ewc_loss_diag": 6.293703336268663e-10, "ewc_loss_parallel": 1.4479155652225018e-09, "grad_norm": 4.817846298217773, "learning_rate": 1.526070368800339e-08, "loss": 0.7934, "mean_token_accuracy": 0.7805756330490112, "num_tokens": 1394696.0, "step": 37 }, { "epoch": 0.004833990586439385, "ewc_loss": 1.7136335372924805e-06, "ewc_loss_diag": 7.166818249970675e-10, "ewc_loss_parallel": 1.6370904631912708e-09, "grad_norm": 4.3797807693481445, "learning_rate": 1.5684612123781262e-08, "loss": 0.7827, "mean_token_accuracy": 0.7824397087097168, "num_tokens": 1438738.0, "step": 38 }, { "epoch": 0.004961200865029895, "ewc_loss": 1.8849968910217285e-06, "ewc_loss_diag": 7.639755494892597e-10, "ewc_loss_parallel": 1.811713445931673e-09, "grad_norm": 4.763704776763916, "learning_rate": 1.6108520559559135e-08, "loss": 0.7482, "mean_token_accuracy": 0.792400598526001, "num_tokens": 1475089.0, "step": 39 }, { "epoch": 0.005088411143620405, "ewc_loss": 1.996755599975586e-06, "ewc_loss_diag": 8.076312951743603e-10, "ewc_loss_parallel": 1.9208528101444244e-09, "grad_norm": 4.747236251831055, "learning_rate": 1.6532428995337004e-08, "loss": 0.808, "mean_token_accuracy": 0.7747877836227417, "num_tokens": 1514566.0, "step": 40 }, { "epoch": 0.005215621422210915, "ewc_loss": 2.0563602447509766e-06, "ewc_loss_diag": 8.585629984736443e-10, "ewc_loss_parallel": 1.964508555829525e-09, "grad_norm": 4.958962440490723, "learning_rate": 1.6956337431114877e-08, "loss": 0.8232, "mean_token_accuracy": 0.7733151316642761, "num_tokens": 1552560.0, "step": 41 }, { "epoch": 0.005342831700801425, "ewc_loss": 2.115964889526367e-06, "ewc_loss_diag": 8.767528925091028e-10, "ewc_loss_parallel": 2.0227162167429924e-09, "grad_norm": 5.386231899261475, "learning_rate": 1.738024586689275e-08, "loss": 0.8006, "mean_token_accuracy": 0.7769511938095093, "num_tokens": 1584759.0, "step": 42 }, { "epoch": 0.005470041979391935, "ewc_loss": 2.175569534301758e-06, "ewc_loss_diag": 9.531504474580288e-10, "ewc_loss_parallel": 2.08092387765646e-09, "grad_norm": 4.851687908172607, "learning_rate": 1.7804154302670622e-08, "loss": 0.8021, "mean_token_accuracy": 0.7759620547294617, "num_tokens": 1621825.0, "step": 43 }, { "epoch": 0.005597252257982445, "ewc_loss": 2.3245811462402344e-06, "ewc_loss_diag": 1.04046193882823e-09, "ewc_loss_parallel": 2.2264430299401283e-09, "grad_norm": 4.783711910247803, "learning_rate": 1.8228062738448494e-08, "loss": 0.8278, "mean_token_accuracy": 0.7682397961616516, "num_tokens": 1662946.0, "step": 44 }, { "epoch": 0.005724462536572955, "ewc_loss": 2.473592758178711e-06, "ewc_loss_diag": 1.1059455573558807e-09, "ewc_loss_parallel": 2.371962182223797e-09, "grad_norm": 4.638556003570557, "learning_rate": 1.8651971174226364e-08, "loss": 0.8153, "mean_token_accuracy": 0.7732576131820679, "num_tokens": 1699433.0, "step": 45 }, { "epoch": 0.005851672815163465, "ewc_loss": 2.5779008865356445e-06, "ewc_loss_diag": 1.1568772606551647e-09, "ewc_loss_parallel": 2.459273673593998e-09, "grad_norm": 4.341951847076416, "learning_rate": 1.9075879610004236e-08, "loss": 0.7463, "mean_token_accuracy": 0.7888889312744141, "num_tokens": 1742812.0, "step": 46 }, { "epoch": 0.005978883093753975, "ewc_loss": 2.682209014892578e-06, "ewc_loss_diag": 1.2078089639544487e-09, "ewc_loss_parallel": 2.561137080192566e-09, "grad_norm": 4.846816062927246, "learning_rate": 1.949978804578211e-08, "loss": 0.8014, "mean_token_accuracy": 0.7783594131469727, "num_tokens": 1778725.0, "step": 47 }, { "epoch": 0.006106093372344485, "ewc_loss": 2.9206275939941406e-06, "ewc_loss_diag": 1.2514647096395493e-09, "ewc_loss_parallel": 2.8085196390748024e-09, "grad_norm": 4.818389415740967, "learning_rate": 1.9923696481559985e-08, "loss": 0.8477, "mean_token_accuracy": 0.7620888352394104, "num_tokens": 1816592.0, "step": 48 }, { "epoch": 0.006233303650934996, "ewc_loss": 3.844499588012695e-06, "ewc_loss_diag": 1.724401954561472e-09, "ewc_loss_parallel": 3.6816345527768135e-09, "grad_norm": 4.382584571838379, "learning_rate": 2.0347604917337857e-08, "loss": 0.7396, "mean_token_accuracy": 0.7896479368209839, "num_tokens": 1859907.0, "step": 49 }, { "epoch": 0.006360513929525506, "ewc_loss": 6.0498714447021484e-06, "ewc_loss_diag": 3.332388587296009e-09, "ewc_loss_parallel": 5.73345459997654e-09, "grad_norm": 4.644885063171387, "learning_rate": 2.0771513353115727e-08, "loss": 0.7289, "mean_token_accuracy": 0.7995734214782715, "num_tokens": 1896627.0, "step": 50 }, { "epoch": 0.006487724208116016, "ewc_loss": 7.987022399902344e-06, "ewc_loss_diag": 4.336470738053322e-09, "ewc_loss_parallel": 7.566995918750763e-09, "grad_norm": 5.077759265899658, "learning_rate": 2.11954217888936e-08, "loss": 0.8444, "mean_token_accuracy": 0.7669410705566406, "num_tokens": 1934041.0, "step": 51 }, { "epoch": 0.006614934486706526, "ewc_loss": 9.298324584960938e-06, "ewc_loss_diag": 4.802132025361061e-09, "ewc_loss_parallel": 8.847564458847046e-09, "grad_norm": 4.561229228973389, "learning_rate": 2.1619330224671472e-08, "loss": 0.7737, "mean_token_accuracy": 0.7840933799743652, "num_tokens": 1976482.0, "step": 52 }, { "epoch": 0.006742144765297036, "ewc_loss": 1.0251998901367188e-05, "ewc_loss_diag": 5.093170329928398e-09, "ewc_loss_parallel": 9.720679372549057e-09, "grad_norm": 5.288649559020996, "learning_rate": 2.2043238660449344e-08, "loss": 0.8031, "mean_token_accuracy": 0.77222740650177, "num_tokens": 2009224.0, "step": 53 }, { "epoch": 0.006869355043887546, "ewc_loss": 1.0967254638671875e-05, "ewc_loss_diag": 5.3551048040390015e-09, "ewc_loss_parallel": 1.0419171303510666e-08, "grad_norm": 5.008654594421387, "learning_rate": 2.2467147096227214e-08, "loss": 0.8473, "mean_token_accuracy": 0.7620739936828613, "num_tokens": 2049235.0, "step": 54 }, { "epoch": 0.006996565322478056, "ewc_loss": 1.1622905731201172e-05, "ewc_loss_diag": 5.617039278149605e-09, "ewc_loss_parallel": 1.1059455573558807e-08, "grad_norm": 4.606873989105225, "learning_rate": 2.2891055532005086e-08, "loss": 0.8603, "mean_token_accuracy": 0.761462926864624, "num_tokens": 2090260.0, "step": 55 }, { "epoch": 0.007123775601068566, "ewc_loss": 1.239776611328125e-05, "ewc_loss_diag": 5.9371814131736755e-09, "ewc_loss_parallel": 1.1816155165433884e-08, "grad_norm": 4.9998931884765625, "learning_rate": 2.331496396778296e-08, "loss": 0.8659, "mean_token_accuracy": 0.7608352899551392, "num_tokens": 2126686.0, "step": 56 }, { "epoch": 0.007250985879659076, "ewc_loss": 1.2755393981933594e-05, "ewc_loss_diag": 6.05359673500061e-09, "ewc_loss_parallel": 1.2165401130914688e-08, "grad_norm": 4.35085391998291, "learning_rate": 2.373887240356083e-08, "loss": 0.7677, "mean_token_accuracy": 0.784743070602417, "num_tokens": 2171355.0, "step": 57 }, { "epoch": 0.007378196158249586, "ewc_loss": 1.3113021850585938e-05, "ewc_loss_diag": 6.28642737865448e-09, "ewc_loss_parallel": 1.2456439435482025e-08, "grad_norm": 4.608287811279297, "learning_rate": 2.4162780839338704e-08, "loss": 0.7409, "mean_token_accuracy": 0.7927963733673096, "num_tokens": 2211660.0, "step": 58 }, { "epoch": 0.007505406436840096, "ewc_loss": 1.329183578491211e-05, "ewc_loss_diag": 6.461050361394882e-09, "ewc_loss_parallel": 1.2631062418222427e-08, "grad_norm": 5.504604816436768, "learning_rate": 2.4586689275116573e-08, "loss": 0.8709, "mean_token_accuracy": 0.7595874071121216, "num_tokens": 2244411.0, "step": 59 }, { "epoch": 0.007632616715430607, "ewc_loss": 1.33514404296875e-05, "ewc_loss_diag": 6.868503987789154e-09, "ewc_loss_parallel": 1.2631062418222427e-08, "grad_norm": 4.347451210021973, "learning_rate": 2.5010597710894446e-08, "loss": 0.7976, "mean_token_accuracy": 0.7710309028625488, "num_tokens": 2285253.0, "step": 60 }, { "epoch": 0.007759826994021117, "ewc_loss": 1.3709068298339844e-05, "ewc_loss_diag": 7.450580596923828e-09, "ewc_loss_parallel": 1.2922100722789764e-08, "grad_norm": 4.493499755859375, "learning_rate": 2.5434506146672318e-08, "loss": 0.7621, "mean_token_accuracy": 0.7854140996932983, "num_tokens": 2328577.0, "step": 61 }, { "epoch": 0.007887037272611627, "ewc_loss": 1.4007091522216797e-05, "ewc_loss_diag": 7.799826562404633e-09, "ewc_loss_parallel": 1.3213139027357101e-08, "grad_norm": 4.688358306884766, "learning_rate": 2.585841458245019e-08, "loss": 0.8422, "mean_token_accuracy": 0.7696568965911865, "num_tokens": 2366524.0, "step": 62 }, { "epoch": 0.008014247551202136, "ewc_loss": 1.4483928680419922e-05, "ewc_loss_diag": 8.149072527885437e-09, "ewc_loss_parallel": 1.367880031466484e-08, "grad_norm": 4.2009124755859375, "learning_rate": 2.628232301822806e-08, "loss": 0.7556, "mean_token_accuracy": 0.7893829345703125, "num_tokens": 2408628.0, "step": 63 }, { "epoch": 0.008141457829792647, "ewc_loss": 1.4543533325195312e-05, "ewc_loss_diag": 8.381903171539307e-09, "ewc_loss_parallel": 1.3737007975578308e-08, "grad_norm": 4.278009414672852, "learning_rate": 2.6706231454005933e-08, "loss": 0.7949, "mean_token_accuracy": 0.7754979133605957, "num_tokens": 2451800.0, "step": 64 }, { "epoch": 0.008268668108383158, "ewc_loss": 1.4722347259521484e-05, "ewc_loss_diag": 8.731149137020111e-09, "ewc_loss_parallel": 1.3853423297405243e-08, "grad_norm": 5.658596515655518, "learning_rate": 2.7130139889783805e-08, "loss": 0.8661, "mean_token_accuracy": 0.75537109375, "num_tokens": 2481448.0, "step": 65 }, { "epoch": 0.008395878386973667, "ewc_loss": 1.519918441772461e-05, "ewc_loss_diag": 9.022187441587448e-09, "ewc_loss_parallel": 1.4319084584712982e-08, "grad_norm": 4.261582851409912, "learning_rate": 2.7554048325561678e-08, "loss": 0.7928, "mean_token_accuracy": 0.775269091129303, "num_tokens": 2526339.0, "step": 66 }, { "epoch": 0.008523088665564178, "ewc_loss": 1.5974044799804688e-05, "ewc_loss_diag": 9.255018085241318e-09, "ewc_loss_parallel": 1.501757651567459e-08, "grad_norm": 4.424585342407227, "learning_rate": 2.797795676133955e-08, "loss": 0.8178, "mean_token_accuracy": 0.770636796951294, "num_tokens": 2570691.0, "step": 67 }, { "epoch": 0.008650298944154687, "ewc_loss": 1.71661376953125e-05, "ewc_loss_diag": 9.778887033462524e-09, "ewc_loss_parallel": 1.618172973394394e-08, "grad_norm": 4.567164897918701, "learning_rate": 2.840186519711742e-08, "loss": 0.8045, "mean_token_accuracy": 0.7751945853233337, "num_tokens": 2609207.0, "step": 68 }, { "epoch": 0.008777509222745198, "ewc_loss": 2.3603439331054688e-05, "ewc_loss_diag": 1.4202669262886047e-08, "ewc_loss_parallel": 2.2118911147117615e-08, "grad_norm": 4.902427673339844, "learning_rate": 2.8825773632895292e-08, "loss": 0.7924, "mean_token_accuracy": 0.7775553464889526, "num_tokens": 2645494.0, "step": 69 }, { "epoch": 0.008904719501335707, "ewc_loss": 3.4332275390625e-05, "ewc_loss_diag": 2.0023435354232788e-08, "ewc_loss_parallel": 3.236345946788788e-08, "grad_norm": 4.429805278778076, "learning_rate": 2.9249682068673165e-08, "loss": 0.7499, "mean_token_accuracy": 0.791788637638092, "num_tokens": 2686258.0, "step": 70 }, { "epoch": 0.009031929779926218, "ewc_loss": 4.291534423828125e-05, "ewc_loss_diag": 2.491287887096405e-08, "ewc_loss_parallel": 4.0512531995773315e-08, "grad_norm": 5.021019458770752, "learning_rate": 2.9673590504451037e-08, "loss": 0.8045, "mean_token_accuracy": 0.7779951095581055, "num_tokens": 2720544.0, "step": 71 }, { "epoch": 0.009159140058516728, "ewc_loss": 4.982948303222656e-05, "ewc_loss_diag": 2.7939677238464355e-08, "ewc_loss_parallel": 4.7031790018081665e-08, "grad_norm": 4.8296966552734375, "learning_rate": 3.0097498940228907e-08, "loss": 0.7677, "mean_token_accuracy": 0.7857566475868225, "num_tokens": 2758068.0, "step": 72 }, { "epoch": 0.009286350337107238, "ewc_loss": 5.435943603515625e-05, "ewc_loss_diag": 2.9569491744041443e-08, "ewc_loss_parallel": 5.145557224750519e-08, "grad_norm": 5.040033340454102, "learning_rate": 3.052140737600678e-08, "loss": 0.8515, "mean_token_accuracy": 0.762520968914032, "num_tokens": 2793342.0, "step": 73 }, { "epoch": 0.009413560615697748, "ewc_loss": 5.841255187988281e-05, "ewc_loss_diag": 3.073364496231079e-08, "ewc_loss_parallel": 5.541369318962097e-08, "grad_norm": 5.2005462646484375, "learning_rate": 3.094531581178465e-08, "loss": 0.8884, "mean_token_accuracy": 0.7551467418670654, "num_tokens": 2828003.0, "step": 74 }, { "epoch": 0.009540770894288259, "ewc_loss": 6.103515625e-05, "ewc_loss_diag": 3.166496753692627e-08, "ewc_loss_parallel": 5.797483026981354e-08, "grad_norm": 4.23414421081543, "learning_rate": 3.1369224247562524e-08, "loss": 0.7639, "mean_token_accuracy": 0.7837315797805786, "num_tokens": 2874755.0, "step": 75 }, { "epoch": 0.00966798117287877, "ewc_loss": 6.151199340820312e-05, "ewc_loss_diag": 3.306195139884949e-08, "ewc_loss_parallel": 5.844049155712128e-08, "grad_norm": 4.630612850189209, "learning_rate": 3.17931326833404e-08, "loss": 0.786, "mean_token_accuracy": 0.7766478061676025, "num_tokens": 2909703.0, "step": 76 }, { "epoch": 0.009795191451469279, "ewc_loss": 6.246566772460938e-05, "ewc_loss_diag": 3.4226104617118835e-08, "ewc_loss_parallel": 5.8906152844429016e-08, "grad_norm": 4.149625301361084, "learning_rate": 3.221704111911827e-08, "loss": 0.7163, "mean_token_accuracy": 0.7983556985855103, "num_tokens": 2954020.0, "step": 77 }, { "epoch": 0.00992240173005979, "ewc_loss": 6.198883056640625e-05, "ewc_loss_diag": 3.562308847904205e-08, "ewc_loss_parallel": 5.8673322200775146e-08, "grad_norm": 5.220978260040283, "learning_rate": 3.264094955489614e-08, "loss": 0.8114, "mean_token_accuracy": 0.7751576900482178, "num_tokens": 2984894.0, "step": 78 }, { "epoch": 0.010049612008650299, "ewc_loss": 6.246566772460938e-05, "ewc_loss_diag": 3.748573362827301e-08, "ewc_loss_parallel": 5.8673322200775146e-08, "grad_norm": 4.938945770263672, "learning_rate": 3.306485799067401e-08, "loss": 0.7185, "mean_token_accuracy": 0.7953177094459534, "num_tokens": 3017773.0, "step": 79 }, { "epoch": 0.01017682228724081, "ewc_loss": 6.29425048828125e-05, "ewc_loss_diag": 3.818422555923462e-08, "ewc_loss_parallel": 5.8906152844429016e-08, "grad_norm": 4.269176959991455, "learning_rate": 3.348876642645188e-08, "loss": 0.7177, "mean_token_accuracy": 0.7980255484580994, "num_tokens": 3059739.0, "step": 80 }, { "epoch": 0.010304032565831319, "ewc_loss": 6.246566772460938e-05, "ewc_loss_diag": 3.91155481338501e-08, "ewc_loss_parallel": 5.8673322200775146e-08, "grad_norm": 4.840972900390625, "learning_rate": 3.391267486222975e-08, "loss": 0.8099, "mean_token_accuracy": 0.7718148231506348, "num_tokens": 3094910.0, "step": 81 }, { "epoch": 0.01043124284442183, "ewc_loss": 6.246566772460938e-05, "ewc_loss_diag": 3.958120942115784e-08, "ewc_loss_parallel": 5.8673322200775146e-08, "grad_norm": 4.566930294036865, "learning_rate": 3.4336583298007626e-08, "loss": 0.8136, "mean_token_accuracy": 0.7731706500053406, "num_tokens": 3133884.0, "step": 82 }, { "epoch": 0.010558453123012339, "ewc_loss": 6.246566772460938e-05, "ewc_loss_diag": 4.0512531995773315e-08, "ewc_loss_parallel": 5.844049155712128e-08, "grad_norm": 5.145877838134766, "learning_rate": 3.47604917337855e-08, "loss": 0.8464, "mean_token_accuracy": 0.7610708475112915, "num_tokens": 3166755.0, "step": 83 }, { "epoch": 0.01068566340160285, "ewc_loss": 6.0558319091796875e-05, "ewc_loss_diag": 4.1443854570388794e-08, "ewc_loss_parallel": 5.657784640789032e-08, "grad_norm": 4.724050521850586, "learning_rate": 3.518440016956337e-08, "loss": 0.8149, "mean_token_accuracy": 0.7709312438964844, "num_tokens": 3204825.0, "step": 84 }, { "epoch": 0.010812873680193359, "ewc_loss": 6.031990051269531e-05, "ewc_loss_diag": 4.470348358154297e-08, "ewc_loss_parallel": 5.587935447692871e-08, "grad_norm": 4.722089767456055, "learning_rate": 3.5608308605341244e-08, "loss": 0.8506, "mean_token_accuracy": 0.7624791860580444, "num_tokens": 3244036.0, "step": 85 }, { "epoch": 0.01094008395878387, "ewc_loss": 6.0558319091796875e-05, "ewc_loss_diag": 4.7963112592697144e-08, "ewc_loss_parallel": 5.587935447692871e-08, "grad_norm": 4.476216793060303, "learning_rate": 3.6032217041119116e-08, "loss": 0.8379, "mean_token_accuracy": 0.7671998143196106, "num_tokens": 3285969.0, "step": 86 }, { "epoch": 0.01106729423737438, "ewc_loss": 6.103515625e-05, "ewc_loss_diag": 4.98257577419281e-08, "ewc_loss_parallel": 5.611218512058258e-08, "grad_norm": 4.443127155303955, "learning_rate": 3.645612547689699e-08, "loss": 0.809, "mean_token_accuracy": 0.7744183540344238, "num_tokens": 3327648.0, "step": 87 }, { "epoch": 0.01119450451596489, "ewc_loss": 6.151199340820312e-05, "ewc_loss_diag": 5.192123353481293e-08, "ewc_loss_parallel": 5.657784640789032e-08, "grad_norm": 4.5860795974731445, "learning_rate": 3.6880033912674855e-08, "loss": 0.8122, "mean_token_accuracy": 0.7729302644729614, "num_tokens": 3367399.0, "step": 88 }, { "epoch": 0.0113217147945554, "ewc_loss": 6.198883056640625e-05, "ewc_loss_diag": 5.3318217396736145e-08, "ewc_loss_parallel": 5.657784640789032e-08, "grad_norm": 4.783356666564941, "learning_rate": 3.730394234845273e-08, "loss": 0.7952, "mean_token_accuracy": 0.779994010925293, "num_tokens": 3405402.0, "step": 89 }, { "epoch": 0.01144892507314591, "ewc_loss": 6.341934204101562e-05, "ewc_loss_diag": 5.494803190231323e-08, "ewc_loss_parallel": 5.797483026981354e-08, "grad_norm": 4.89097785949707, "learning_rate": 3.77278507842306e-08, "loss": 0.8789, "mean_token_accuracy": 0.7545250654220581, "num_tokens": 3441791.0, "step": 90 }, { "epoch": 0.01157613535173642, "ewc_loss": 6.437301635742188e-05, "ewc_loss_diag": 5.634501576423645e-08, "ewc_loss_parallel": 5.8673322200775146e-08, "grad_norm": 4.819455146789551, "learning_rate": 3.815175922000847e-08, "loss": 0.8381, "mean_token_accuracy": 0.7690532803535461, "num_tokens": 3480151.0, "step": 91 }, { "epoch": 0.01170334563032693, "ewc_loss": 6.628036499023438e-05, "ewc_loss_diag": 5.820766091346741e-08, "ewc_loss_parallel": 6.05359673500061e-08, "grad_norm": 4.7391486167907715, "learning_rate": 3.8575667655786345e-08, "loss": 0.8019, "mean_token_accuracy": 0.7763080596923828, "num_tokens": 3516867.0, "step": 92 }, { "epoch": 0.01183055590891744, "ewc_loss": 6.866455078125e-05, "ewc_loss_diag": 5.960464477539063e-08, "ewc_loss_parallel": 6.239861249923706e-08, "grad_norm": 4.298402309417725, "learning_rate": 3.899957609156422e-08, "loss": 0.7777, "mean_token_accuracy": 0.7822694778442383, "num_tokens": 3557466.0, "step": 93 }, { "epoch": 0.01195776618750795, "ewc_loss": 7.390975952148438e-05, "ewc_loss_diag": 6.007030606269836e-08, "ewc_loss_parallel": 6.752088665962219e-08, "grad_norm": 4.535126209259033, "learning_rate": 3.94234845273421e-08, "loss": 0.8072, "mean_token_accuracy": 0.773403525352478, "num_tokens": 3596009.0, "step": 94 }, { "epoch": 0.012084976466098461, "ewc_loss": 8.535385131835938e-05, "ewc_loss_diag": 6.146728992462158e-08, "ewc_loss_parallel": 7.916241884231567e-08, "grad_norm": 4.751617431640625, "learning_rate": 3.984739296311997e-08, "loss": 0.7993, "mean_token_accuracy": 0.7744667530059814, "num_tokens": 3631582.0, "step": 95 }, { "epoch": 0.01221218674468897, "ewc_loss": 0.00011682510375976562, "ewc_loss_diag": 6.51925802230835e-08, "ewc_loss_parallel": 1.103617250919342e-07, "grad_norm": 4.72960901260376, "learning_rate": 4.027130139889784e-08, "loss": 0.826, "mean_token_accuracy": 0.7680369019508362, "num_tokens": 3672855.0, "step": 96 }, { "epoch": 0.012339397023279481, "ewc_loss": 0.00016689300537109375, "ewc_loss_diag": 7.683411240577698e-08, "ewc_loss_parallel": 1.5925616025924683e-07, "grad_norm": 4.53072452545166, "learning_rate": 4.0695209834675715e-08, "loss": 0.8176, "mean_token_accuracy": 0.7728264331817627, "num_tokens": 3714100.0, "step": 97 }, { "epoch": 0.012466607301869992, "ewc_loss": 0.00022125244140625, "ewc_loss_diag": 9.033828973770142e-08, "ewc_loss_parallel": 2.123415470123291e-07, "grad_norm": 4.6208014488220215, "learning_rate": 4.111911827045358e-08, "loss": 0.7687, "mean_token_accuracy": 0.785271942615509, "num_tokens": 3752834.0, "step": 98 }, { "epoch": 0.012593817580460501, "ewc_loss": 0.000255584716796875, "ewc_loss_diag": 1.0058283805847168e-07, "ewc_loss_parallel": 2.4586915969848633e-07, "grad_norm": 4.634740352630615, "learning_rate": 4.154302670623145e-08, "loss": 0.7789, "mean_token_accuracy": 0.7812429070472717, "num_tokens": 3791513.0, "step": 99 }, { "epoch": 0.012721027859051012, "ewc_loss": 0.000270843505859375, "ewc_loss_diag": 1.0896474123001099e-07, "ewc_loss_parallel": 2.5890767574310303e-07, "grad_norm": 5.144634246826172, "learning_rate": 4.1966935142009326e-08, "loss": 0.8019, "mean_token_accuracy": 0.7744066715240479, "num_tokens": 3825085.0, "step": 100 }, { "epoch": 0.012848238137641521, "ewc_loss": 0.0002727508544921875, "ewc_loss_diag": 1.1641532182693481e-07, "ewc_loss_parallel": 2.60770320892334e-07, "grad_norm": 4.527338027954102, "learning_rate": 4.23908435777872e-08, "loss": 0.802, "mean_token_accuracy": 0.7769638895988464, "num_tokens": 3866492.0, "step": 101 }, { "epoch": 0.012975448416232032, "ewc_loss": 0.00026702880859375, "ewc_loss_diag": 1.210719347000122e-07, "ewc_loss_parallel": 2.551823854446411e-07, "grad_norm": 4.888991355895996, "learning_rate": 4.281475201356507e-08, "loss": 0.8293, "mean_token_accuracy": 0.7629509568214417, "num_tokens": 3904053.0, "step": 102 }, { "epoch": 0.013102658694822541, "ewc_loss": 0.00026702880859375, "ewc_loss_diag": 1.2759119272232056e-07, "ewc_loss_parallel": 2.5331974029541016e-07, "grad_norm": 4.902311325073242, "learning_rate": 4.3238660449342943e-08, "loss": 0.7376, "mean_token_accuracy": 0.7953550219535828, "num_tokens": 3935118.0, "step": 103 }, { "epoch": 0.013229868973413052, "ewc_loss": 0.00026702880859375, "ewc_loss_diag": 1.3224780559539795e-07, "ewc_loss_parallel": 2.5331974029541016e-07, "grad_norm": 4.425089359283447, "learning_rate": 4.3662568885120816e-08, "loss": 0.7877, "mean_token_accuracy": 0.776278018951416, "num_tokens": 3977899.0, "step": 104 }, { "epoch": 0.013357079252003561, "ewc_loss": 0.000263214111328125, "ewc_loss_diag": 1.3504177331924438e-07, "ewc_loss_parallel": 2.4959444999694824e-07, "grad_norm": 4.434872150421143, "learning_rate": 4.408647732089869e-08, "loss": 0.8079, "mean_token_accuracy": 0.7737969160079956, "num_tokens": 4019416.0, "step": 105 }, { "epoch": 0.013484289530594072, "ewc_loss": 0.000263214111328125, "ewc_loss_diag": 1.387670636177063e-07, "ewc_loss_parallel": 2.4959444999694824e-07, "grad_norm": 4.340202808380127, "learning_rate": 4.451038575667656e-08, "loss": 0.797, "mean_token_accuracy": 0.7731912136077881, "num_tokens": 4063887.0, "step": 106 }, { "epoch": 0.013611499809184581, "ewc_loss": 0.0002613067626953125, "ewc_loss_diag": 1.424923539161682e-07, "ewc_loss_parallel": 2.477318048477173e-07, "grad_norm": 4.506662368774414, "learning_rate": 4.493429419245443e-08, "loss": 0.8279, "mean_token_accuracy": 0.7675923705101013, "num_tokens": 4101309.0, "step": 107 }, { "epoch": 0.013738710087775092, "ewc_loss": 0.0002593994140625, "ewc_loss_diag": 1.4621764421463013e-07, "ewc_loss_parallel": 2.4400651454925537e-07, "grad_norm": 4.6402812004089355, "learning_rate": 4.53582026282323e-08, "loss": 0.812, "mean_token_accuracy": 0.7706453800201416, "num_tokens": 4136669.0, "step": 108 }, { "epoch": 0.013865920366365603, "ewc_loss": 0.0002536773681640625, "ewc_loss_diag": 1.4901161193847656e-07, "ewc_loss_parallel": 2.384185791015625e-07, "grad_norm": 4.166482448577881, "learning_rate": 4.578211106401017e-08, "loss": 0.7431, "mean_token_accuracy": 0.7902011871337891, "num_tokens": 4178583.0, "step": 109 }, { "epoch": 0.013993130644956112, "ewc_loss": 0.000247955322265625, "ewc_loss_diag": 1.5366822481155396e-07, "ewc_loss_parallel": 2.3283064365386963e-07, "grad_norm": 4.17488956451416, "learning_rate": 4.6206019499788045e-08, "loss": 0.7737, "mean_token_accuracy": 0.777603030204773, "num_tokens": 4219552.0, "step": 110 }, { "epoch": 0.014120340923546623, "ewc_loss": 0.000244140625, "ewc_loss_diag": 1.5739351511001587e-07, "ewc_loss_parallel": 2.2910535335540771e-07, "grad_norm": 4.229003429412842, "learning_rate": 4.662992793556592e-08, "loss": 0.7665, "mean_token_accuracy": 0.7853053212165833, "num_tokens": 4264487.0, "step": 111 }, { "epoch": 0.014247551202137132, "ewc_loss": 0.00023555755615234375, "ewc_loss_diag": 1.5925616025924683e-07, "ewc_loss_parallel": 2.1979212760925293e-07, "grad_norm": 4.324753761291504, "learning_rate": 4.705383637134379e-08, "loss": 0.7326, "mean_token_accuracy": 0.7898605465888977, "num_tokens": 4303015.0, "step": 112 }, { "epoch": 0.014374761480727643, "ewc_loss": 0.00022983551025390625, "ewc_loss_diag": 1.6298145055770874e-07, "ewc_loss_parallel": 2.1327286958694458e-07, "grad_norm": 4.265488624572754, "learning_rate": 4.747774480712166e-08, "loss": 0.7874, "mean_token_accuracy": 0.7769029140472412, "num_tokens": 4345446.0, "step": 113 }, { "epoch": 0.014501971759318152, "ewc_loss": 0.0002231597900390625, "ewc_loss_diag": 1.6670674085617065e-07, "ewc_loss_parallel": 2.0675361156463623e-07, "grad_norm": 4.379148960113525, "learning_rate": 4.7901653242899535e-08, "loss": 0.7398, "mean_token_accuracy": 0.7912427186965942, "num_tokens": 4383854.0, "step": 114 }, { "epoch": 0.014629182037908663, "ewc_loss": 0.00022125244140625, "ewc_loss_diag": 1.7043203115463257e-07, "ewc_loss_parallel": 2.039596438407898e-07, "grad_norm": 4.446916580200195, "learning_rate": 4.832556167867741e-08, "loss": 0.8087, "mean_token_accuracy": 0.7728708386421204, "num_tokens": 4420750.0, "step": 115 }, { "epoch": 0.014756392316499172, "ewc_loss": 0.00021457672119140625, "ewc_loss_diag": 1.7415732145309448e-07, "ewc_loss_parallel": 1.9744038581848145e-07, "grad_norm": 4.572436332702637, "learning_rate": 4.8749470114455274e-08, "loss": 0.8058, "mean_token_accuracy": 0.7730389833450317, "num_tokens": 4458761.0, "step": 116 }, { "epoch": 0.014883602595089683, "ewc_loss": 0.0002079010009765625, "ewc_loss_diag": 1.7695128917694092e-07, "ewc_loss_parallel": 1.909211277961731e-07, "grad_norm": 4.3450846672058105, "learning_rate": 4.9173378550233146e-08, "loss": 0.766, "mean_token_accuracy": 0.7807909846305847, "num_tokens": 4496547.0, "step": 117 }, { "epoch": 0.015010812873680193, "ewc_loss": 0.0002040863037109375, "ewc_loss_diag": 1.8067657947540283e-07, "ewc_loss_parallel": 1.862645149230957e-07, "grad_norm": 4.393194675445557, "learning_rate": 4.959728698601102e-08, "loss": 0.764, "mean_token_accuracy": 0.7898052930831909, "num_tokens": 4533357.0, "step": 118 }, { "epoch": 0.015138023152270703, "ewc_loss": 0.00020313262939453125, "ewc_loss_diag": 1.8719583749771118e-07, "ewc_loss_parallel": 1.8440186977386475e-07, "grad_norm": 4.216264724731445, "learning_rate": 5.002119542178889e-08, "loss": 0.7385, "mean_token_accuracy": 0.7907528281211853, "num_tokens": 4573570.0, "step": 119 }, { "epoch": 0.015265233430861214, "ewc_loss": 0.0002002716064453125, "ewc_loss_diag": 1.9371509552001953e-07, "ewc_loss_parallel": 1.816079020500183e-07, "grad_norm": 4.428993225097656, "learning_rate": 5.0445103857566764e-08, "loss": 0.7937, "mean_token_accuracy": 0.7747960686683655, "num_tokens": 4613195.0, "step": 120 }, { "epoch": 0.015392443709451724, "ewc_loss": 0.00019931793212890625, "ewc_loss_diag": 2.0116567611694336e-07, "ewc_loss_parallel": 1.7881393432617188e-07, "grad_norm": 4.057980537414551, "learning_rate": 5.0869012293344637e-08, "loss": 0.7139, "mean_token_accuracy": 0.7978157997131348, "num_tokens": 4657230.0, "step": 121 }, { "epoch": 0.015519653988042234, "ewc_loss": 0.00019741058349609375, "ewc_loss_diag": 2.1047890186309814e-07, "ewc_loss_parallel": 1.7601996660232544e-07, "grad_norm": 4.2058916091918945, "learning_rate": 5.129292072912251e-08, "loss": 0.7876, "mean_token_accuracy": 0.7810484170913696, "num_tokens": 4701436.0, "step": 122 }, { "epoch": 0.015646864266632744, "ewc_loss": 0.0001964569091796875, "ewc_loss_diag": 2.1886080503463745e-07, "ewc_loss_parallel": 1.7415732145309448e-07, "grad_norm": 4.303966045379639, "learning_rate": 5.171682916490038e-08, "loss": 0.7756, "mean_token_accuracy": 0.780083179473877, "num_tokens": 4741806.0, "step": 123 }, { "epoch": 0.015774074545223254, "ewc_loss": 0.00019550323486328125, "ewc_loss_diag": 2.2631138563156128e-07, "ewc_loss_parallel": 1.7229467630386353e-07, "grad_norm": 4.50932502746582, "learning_rate": 5.2140737600678254e-08, "loss": 0.8124, "mean_token_accuracy": 0.7655160427093506, "num_tokens": 4780671.0, "step": 124 }, { "epoch": 0.015901284823813765, "ewc_loss": 0.00019550323486328125, "ewc_loss_diag": 2.3562461137771606e-07, "ewc_loss_parallel": 1.7136335372924805e-07, "grad_norm": 4.472337245941162, "learning_rate": 5.256464603645612e-08, "loss": 0.7613, "mean_token_accuracy": 0.786303699016571, "num_tokens": 4816879.0, "step": 125 }, { "epoch": 0.016028495102404273, "ewc_loss": 0.00019550323486328125, "ewc_loss_diag": 2.384185791015625e-07, "ewc_loss_parallel": 1.7136335372924805e-07, "grad_norm": 4.369726657867432, "learning_rate": 5.298855447223399e-08, "loss": 0.8135, "mean_token_accuracy": 0.7676681876182556, "num_tokens": 4858704.0, "step": 126 }, { "epoch": 0.016155705380994784, "ewc_loss": 0.0001964569091796875, "ewc_loss_diag": 2.4400651454925537e-07, "ewc_loss_parallel": 1.7136335372924805e-07, "grad_norm": 4.326704025268555, "learning_rate": 5.3412462908011865e-08, "loss": 0.7707, "mean_token_accuracy": 0.782097578048706, "num_tokens": 4896780.0, "step": 127 }, { "epoch": 0.016282915659585295, "ewc_loss": 0.0002002716064453125, "ewc_loss_diag": 2.5331974029541016e-07, "ewc_loss_parallel": 1.7415732145309448e-07, "grad_norm": 4.56742525100708, "learning_rate": 5.383637134378974e-08, "loss": 0.8005, "mean_token_accuracy": 0.7717974185943604, "num_tokens": 4932108.0, "step": 128 }, { "epoch": 0.016410125938175806, "ewc_loss": 0.00020694732666015625, "ewc_loss_diag": 2.5704503059387207e-07, "ewc_loss_parallel": 1.816079020500183e-07, "grad_norm": 4.3037848472595215, "learning_rate": 5.426027977956761e-08, "loss": 0.7768, "mean_token_accuracy": 0.7822123765945435, "num_tokens": 4970384.0, "step": 129 }, { "epoch": 0.016537336216766316, "ewc_loss": 0.0002193450927734375, "ewc_loss_diag": 2.7008354663848877e-07, "ewc_loss_parallel": 1.9278377294540405e-07, "grad_norm": 4.654110431671143, "learning_rate": 5.468418821534548e-08, "loss": 0.7853, "mean_token_accuracy": 0.7767638564109802, "num_tokens": 5005996.0, "step": 130 }, { "epoch": 0.016664546495356824, "ewc_loss": 0.00023937225341796875, "ewc_loss_diag": 2.7567148208618164e-07, "ewc_loss_parallel": 2.1141022443771362e-07, "grad_norm": 4.440157413482666, "learning_rate": 5.5108096651123356e-08, "loss": 0.7649, "mean_token_accuracy": 0.7814390063285828, "num_tokens": 5044208.0, "step": 131 }, { "epoch": 0.016791756773947335, "ewc_loss": 0.000270843505859375, "ewc_loss_diag": 2.8312206268310547e-07, "ewc_loss_parallel": 2.421438694000244e-07, "grad_norm": 4.327836513519287, "learning_rate": 5.553200508690123e-08, "loss": 0.7799, "mean_token_accuracy": 0.7761330008506775, "num_tokens": 5085173.0, "step": 132 }, { "epoch": 0.016918967052537846, "ewc_loss": 0.000316619873046875, "ewc_loss_diag": 2.9243528842926025e-07, "ewc_loss_parallel": 2.8870999813079834e-07, "grad_norm": 4.772838115692139, "learning_rate": 5.59559135226791e-08, "loss": 0.796, "mean_token_accuracy": 0.7739336490631104, "num_tokens": 5118440.0, "step": 133 }, { "epoch": 0.017046177331128357, "ewc_loss": 0.0003871917724609375, "ewc_loss_diag": 2.998858690261841e-07, "ewc_loss_parallel": 3.557652235031128e-07, "grad_norm": 4.859479904174805, "learning_rate": 5.637982195845697e-08, "loss": 0.7999, "mean_token_accuracy": 0.7736718058586121, "num_tokens": 5155202.0, "step": 134 }, { "epoch": 0.017173387609718864, "ewc_loss": 0.0004749298095703125, "ewc_loss_diag": 3.129243850708008e-07, "ewc_loss_parallel": 4.4517219066619873e-07, "grad_norm": 4.747000217437744, "learning_rate": 5.680373039423484e-08, "loss": 0.7853, "mean_token_accuracy": 0.7794568538665771, "num_tokens": 5193240.0, "step": 135 }, { "epoch": 0.017300597888309375, "ewc_loss": 0.000568389892578125, "ewc_loss_diag": 3.259629011154175e-07, "ewc_loss_parallel": 5.327165126800537e-07, "grad_norm": 4.856104850769043, "learning_rate": 5.722763883001271e-08, "loss": 0.7731, "mean_token_accuracy": 0.7807317972183228, "num_tokens": 5230975.0, "step": 136 }, { "epoch": 0.017427808166899886, "ewc_loss": 0.000659942626953125, "ewc_loss_diag": 3.501772880554199e-07, "ewc_loss_parallel": 6.258487701416016e-07, "grad_norm": 5.300549030303955, "learning_rate": 5.7651547265790585e-08, "loss": 0.8347, "mean_token_accuracy": 0.7636812329292297, "num_tokens": 5264786.0, "step": 137 }, { "epoch": 0.017555018445490397, "ewc_loss": 0.00074005126953125, "ewc_loss_diag": 3.594905138015747e-07, "ewc_loss_parallel": 7.040798664093018e-07, "grad_norm": 4.903783798217773, "learning_rate": 5.807545570156846e-08, "loss": 0.7968, "mean_token_accuracy": 0.7744324207305908, "num_tokens": 5301554.0, "step": 138 }, { "epoch": 0.017682228724080904, "ewc_loss": 0.0007781982421875, "ewc_loss_diag": 3.725290298461914e-07, "ewc_loss_parallel": 7.413327693939209e-07, "grad_norm": 4.779110908508301, "learning_rate": 5.849936413734633e-08, "loss": 0.7354, "mean_token_accuracy": 0.7902576923370361, "num_tokens": 5337356.0, "step": 139 }, { "epoch": 0.017809439002671415, "ewc_loss": 0.000820159912109375, "ewc_loss_diag": 3.818422555923462e-07, "ewc_loss_parallel": 7.82310962677002e-07, "grad_norm": 4.610597610473633, "learning_rate": 5.89232725731242e-08, "loss": 0.8067, "mean_token_accuracy": 0.7737829089164734, "num_tokens": 5375767.0, "step": 140 }, { "epoch": 0.017936649281261926, "ewc_loss": 0.00084686279296875, "ewc_loss_diag": 4.0605664253234863e-07, "ewc_loss_parallel": 8.046627044677734e-07, "grad_norm": 4.548935413360596, "learning_rate": 5.9347181008902075e-08, "loss": 0.7572, "mean_token_accuracy": 0.7831001877784729, "num_tokens": 5413683.0, "step": 141 }, { "epoch": 0.018063859559852437, "ewc_loss": 0.0008544921875, "ewc_loss_diag": 4.246830940246582e-07, "ewc_loss_parallel": 8.121132850646973e-07, "grad_norm": 4.2777838706970215, "learning_rate": 5.977108944467995e-08, "loss": 0.7386, "mean_token_accuracy": 0.7884997129440308, "num_tokens": 5457750.0, "step": 142 }, { "epoch": 0.018191069838442948, "ewc_loss": 0.0008544921875, "ewc_loss_diag": 4.4517219066619873e-07, "ewc_loss_parallel": 8.083879947662354e-07, "grad_norm": 4.527960300445557, "learning_rate": 6.019499788045781e-08, "loss": 0.7331, "mean_token_accuracy": 0.7897542119026184, "num_tokens": 5495743.0, "step": 143 }, { "epoch": 0.018318280117033455, "ewc_loss": 0.000850677490234375, "ewc_loss_diag": 4.5821070671081543e-07, "ewc_loss_parallel": 8.046627044677734e-07, "grad_norm": 4.532920837402344, "learning_rate": 6.061890631623569e-08, "loss": 0.7465, "mean_token_accuracy": 0.7815498113632202, "num_tokens": 5532198.0, "step": 144 }, { "epoch": 0.018445490395623966, "ewc_loss": 0.00084686279296875, "ewc_loss_diag": 4.6566128730773926e-07, "ewc_loss_parallel": 8.009374141693115e-07, "grad_norm": 4.689803123474121, "learning_rate": 6.104281475201356e-08, "loss": 0.7587, "mean_token_accuracy": 0.7829457521438599, "num_tokens": 5568977.0, "step": 145 }, { "epoch": 0.018572700674214477, "ewc_loss": 0.0008544921875, "ewc_loss_diag": 4.76837158203125e-07, "ewc_loss_parallel": 8.046627044677734e-07, "grad_norm": 4.690504550933838, "learning_rate": 6.146672318779143e-08, "loss": 0.7714, "mean_token_accuracy": 0.7734643816947937, "num_tokens": 5606229.0, "step": 146 }, { "epoch": 0.018699910952804988, "ewc_loss": 0.0008544921875, "ewc_loss_diag": 4.842877388000488e-07, "ewc_loss_parallel": 8.046627044677734e-07, "grad_norm": 4.411471843719482, "learning_rate": 6.18906316235693e-08, "loss": 0.7724, "mean_token_accuracy": 0.7763131856918335, "num_tokens": 5649828.0, "step": 147 }, { "epoch": 0.018827121231395495, "ewc_loss": 0.000835418701171875, "ewc_loss_diag": 4.880130290985107e-07, "ewc_loss_parallel": 7.860362529754639e-07, "grad_norm": 4.5872039794921875, "learning_rate": 6.231454005934718e-08, "loss": 0.7186, "mean_token_accuracy": 0.7963597774505615, "num_tokens": 5687433.0, "step": 148 }, { "epoch": 0.018954331509986006, "ewc_loss": 0.000823974609375, "ewc_loss_diag": 4.954636096954346e-07, "ewc_loss_parallel": 7.748603820800781e-07, "grad_norm": 4.619120121002197, "learning_rate": 6.273844849512505e-08, "loss": 0.7496, "mean_token_accuracy": 0.7831704616546631, "num_tokens": 5723247.0, "step": 149 }, { "epoch": 0.019081541788576517, "ewc_loss": 0.000812530517578125, "ewc_loss_diag": 5.029141902923584e-07, "ewc_loss_parallel": 7.636845111846924e-07, "grad_norm": 4.430767059326172, "learning_rate": 6.316235693090292e-08, "loss": 0.7643, "mean_token_accuracy": 0.7787022590637207, "num_tokens": 5760305.0, "step": 150 }, { "epoch": 0.019208752067167028, "ewc_loss": 0.000797271728515625, "ewc_loss_diag": 5.103647708892822e-07, "ewc_loss_parallel": 7.487833499908447e-07, "grad_norm": 4.210988521575928, "learning_rate": 6.35862653666808e-08, "loss": 0.7646, "mean_token_accuracy": 0.7770370244979858, "num_tokens": 5800586.0, "step": 151 }, { "epoch": 0.01933596234575754, "ewc_loss": 0.00077056884765625, "ewc_loss_diag": 5.140900611877441e-07, "ewc_loss_parallel": 7.227063179016113e-07, "grad_norm": 4.1913042068481445, "learning_rate": 6.401017380245867e-08, "loss": 0.7561, "mean_token_accuracy": 0.7791637182235718, "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": 4.247735977172852, "learning_rate": 6.443408223823654e-08, "loss": 0.7966, "mean_token_accuracy": 0.7701898813247681, "num_tokens": 5880162.0, "step": 153 }, { "epoch": 0.019590382902938557, "ewc_loss": 0.00074005126953125, "ewc_loss_diag": 5.364418029785156e-07, "ewc_loss_parallel": 6.854534149169922e-07, "grad_norm": 4.319072246551514, "learning_rate": 6.485799067401441e-08, "loss": 0.8522, "mean_token_accuracy": 0.7552526593208313, "num_tokens": 5922213.0, "step": 154 }, { "epoch": 0.019717593181529068, "ewc_loss": 0.000728607177734375, "ewc_loss_diag": 5.438923835754395e-07, "ewc_loss_parallel": 6.742775440216064e-07, "grad_norm": 4.238588809967041, "learning_rate": 6.528189910979228e-08, "loss": 0.734, "mean_token_accuracy": 0.7858923673629761, "num_tokens": 5957461.0, "step": 155 }, { "epoch": 0.01984480346011958, "ewc_loss": 0.000713348388671875, "ewc_loss_diag": 5.513429641723633e-07, "ewc_loss_parallel": 6.593763828277588e-07, "grad_norm": 4.159764766693115, "learning_rate": 6.570580754557016e-08, "loss": 0.7194, "mean_token_accuracy": 0.7857296466827393, "num_tokens": 5994675.0, "step": 156 }, { "epoch": 0.019972013738710086, "ewc_loss": 0.000701904296875, "ewc_loss_diag": 5.587935447692871e-07, "ewc_loss_parallel": 6.444752216339111e-07, "grad_norm": 4.179564952850342, "learning_rate": 6.612971598134802e-08, "loss": 0.6821, "mean_token_accuracy": 0.8014270663261414, "num_tokens": 6032514.0, "step": 157 }, { "epoch": 0.020099224017300597, "ewc_loss": 0.00067901611328125, "ewc_loss_diag": 5.62518835067749e-07, "ewc_loss_parallel": 6.221234798431396e-07, "grad_norm": 4.195849895477295, "learning_rate": 6.655362441712589e-08, "loss": 0.7543, "mean_token_accuracy": 0.7800713777542114, "num_tokens": 6069160.0, "step": 158 }, { "epoch": 0.020226434295891108, "ewc_loss": 0.000667572021484375, "ewc_loss_diag": 5.662441253662109e-07, "ewc_loss_parallel": 6.109476089477539e-07, "grad_norm": 4.3348259925842285, "learning_rate": 6.697753285290376e-08, "loss": 0.8086, "mean_token_accuracy": 0.7646182179450989, "num_tokens": 6106652.0, "step": 159 }, { "epoch": 0.02035364457448162, "ewc_loss": 0.0006561279296875, "ewc_loss_diag": 5.811452865600586e-07, "ewc_loss_parallel": 5.997717380523682e-07, "grad_norm": 4.121046543121338, "learning_rate": 6.740144128868163e-08, "loss": 0.7429, "mean_token_accuracy": 0.7862206697463989, "num_tokens": 6143070.0, "step": 160 }, { "epoch": 0.020480854853072127, "ewc_loss": 0.00063323974609375, "ewc_loss_diag": 5.923211574554443e-07, "ewc_loss_parallel": 5.736947059631348e-07, "grad_norm": 3.855644464492798, "learning_rate": 6.78253497244595e-08, "loss": 0.7351, "mean_token_accuracy": 0.7867216467857361, "num_tokens": 6186495.0, "step": 161 }, { "epoch": 0.020608065131662637, "ewc_loss": 0.000621795654296875, "ewc_loss_diag": 6.07222318649292e-07, "ewc_loss_parallel": 5.62518835067749e-07, "grad_norm": 4.221208095550537, "learning_rate": 6.824925816023738e-08, "loss": 0.7359, "mean_token_accuracy": 0.7861763834953308, "num_tokens": 6222454.0, "step": 162 }, { "epoch": 0.02073527541025315, "ewc_loss": 0.0006103515625, "ewc_loss_diag": 6.183981895446777e-07, "ewc_loss_parallel": 5.476176738739014e-07, "grad_norm": 4.447734355926514, "learning_rate": 6.867316659601525e-08, "loss": 0.7484, "mean_token_accuracy": 0.7843630313873291, "num_tokens": 6253260.0, "step": 163 }, { "epoch": 0.02086248568884366, "ewc_loss": 0.0006103515625, "ewc_loss_diag": 6.444752216339111e-07, "ewc_loss_parallel": 5.438923835754395e-07, "grad_norm": 3.907132387161255, "learning_rate": 6.909707503179312e-08, "loss": 0.7291, "mean_token_accuracy": 0.7901371717453003, "num_tokens": 6294773.0, "step": 164 }, { "epoch": 0.02098969596743417, "ewc_loss": 0.00060272216796875, "ewc_loss_diag": 6.51925802230835e-07, "ewc_loss_parallel": 5.364418029785156e-07, "grad_norm": 3.967755079269409, "learning_rate": 6.9520983467571e-08, "loss": 0.7727, "mean_token_accuracy": 0.7758862972259521, "num_tokens": 6335334.0, "step": 165 }, { "epoch": 0.021116906246024678, "ewc_loss": 0.0005950927734375, "ewc_loss_diag": 6.631016731262207e-07, "ewc_loss_parallel": 5.289912223815918e-07, "grad_norm": 3.9553661346435547, "learning_rate": 6.994489190334887e-08, "loss": 0.7077, "mean_token_accuracy": 0.7914961576461792, "num_tokens": 6372255.0, "step": 166 }, { "epoch": 0.02124411652461519, "ewc_loss": 0.000591278076171875, "ewc_loss_diag": 6.705522537231445e-07, "ewc_loss_parallel": 5.21540641784668e-07, "grad_norm": 3.8210246562957764, "learning_rate": 7.036880033912674e-08, "loss": 0.6842, "mean_token_accuracy": 0.8004199862480164, "num_tokens": 6411971.0, "step": 167 }, { "epoch": 0.0213713268032057, "ewc_loss": 0.000576019287109375, "ewc_loss_diag": 6.817281246185303e-07, "ewc_loss_parallel": 5.066394805908203e-07, "grad_norm": 4.177430152893066, "learning_rate": 7.079270877490461e-08, "loss": 0.7546, "mean_token_accuracy": 0.7795610427856445, "num_tokens": 6450366.0, "step": 168 }, { "epoch": 0.02149853708179621, "ewc_loss": 0.00057220458984375, "ewc_loss_diag": 6.966292858123779e-07, "ewc_loss_parallel": 5.029141902923584e-07, "grad_norm": 4.213230133056641, "learning_rate": 7.121661721068249e-08, "loss": 0.7172, "mean_token_accuracy": 0.7890644073486328, "num_tokens": 6482199.0, "step": 169 }, { "epoch": 0.021625747360386718, "ewc_loss": 0.00057220458984375, "ewc_loss_diag": 7.078051567077637e-07, "ewc_loss_parallel": 5.029141902923584e-07, "grad_norm": 3.894662380218506, "learning_rate": 7.164052564646036e-08, "loss": 0.6989, "mean_token_accuracy": 0.7982782125473022, "num_tokens": 6519697.0, "step": 170 }, { "epoch": 0.02175295763897723, "ewc_loss": 0.000568389892578125, "ewc_loss_diag": 7.189810276031494e-07, "ewc_loss_parallel": 4.954636096954346e-07, "grad_norm": 4.037126541137695, "learning_rate": 7.206443408223823e-08, "loss": 0.7594, "mean_token_accuracy": 0.778531551361084, "num_tokens": 6554687.0, "step": 171 }, { "epoch": 0.02188016791756774, "ewc_loss": 0.000560760498046875, "ewc_loss_diag": 7.264316082000732e-07, "ewc_loss_parallel": 4.880130290985107e-07, "grad_norm": 4.032703399658203, "learning_rate": 7.24883425180161e-08, "loss": 0.7222, "mean_token_accuracy": 0.7828387022018433, "num_tokens": 6593066.0, "step": 172 }, { "epoch": 0.02200737819615825, "ewc_loss": 0.00055694580078125, "ewc_loss_diag": 7.450580596923828e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 3.693859577178955, "learning_rate": 7.291225095379398e-08, "loss": 0.6801, "mean_token_accuracy": 0.8026199340820312, "num_tokens": 6635484.0, "step": 173 }, { "epoch": 0.02213458847474876, "ewc_loss": 0.000560760498046875, "ewc_loss_diag": 7.525086402893066e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 3.7600350379943848, "learning_rate": 7.333615938957185e-08, "loss": 0.7725, "mean_token_accuracy": 0.7759203314781189, "num_tokens": 6677090.0, "step": 174 }, { "epoch": 0.02226179875333927, "ewc_loss": 0.000560760498046875, "ewc_loss_diag": 7.711350917816162e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 4.1220221519470215, "learning_rate": 7.376006782534971e-08, "loss": 0.6876, "mean_token_accuracy": 0.796485185623169, "num_tokens": 6710250.0, "step": 175 }, { "epoch": 0.02238900903192978, "ewc_loss": 0.0005645751953125, "ewc_loss_diag": 7.897615432739258e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 3.744378089904785, "learning_rate": 7.418397626112758e-08, "loss": 0.7141, "mean_token_accuracy": 0.7885410189628601, "num_tokens": 6752998.0, "step": 176 }, { "epoch": 0.02251621931052029, "ewc_loss": 0.000568389892578125, "ewc_loss_diag": 8.009374141693115e-07, "ewc_loss_parallel": 4.880130290985107e-07, "grad_norm": 3.949812889099121, "learning_rate": 7.460788469690545e-08, "loss": 0.7153, "mean_token_accuracy": 0.7945334911346436, "num_tokens": 6789568.0, "step": 177 }, { "epoch": 0.0226434295891108, "ewc_loss": 0.000576019287109375, "ewc_loss_diag": 8.158385753631592e-07, "ewc_loss_parallel": 4.954636096954346e-07, "grad_norm": 4.45575475692749, "learning_rate": 7.503179313268333e-08, "loss": 0.7563, "mean_token_accuracy": 0.7778602838516235, "num_tokens": 6822810.0, "step": 178 }, { "epoch": 0.02277063986770131, "ewc_loss": 0.0005950927734375, "ewc_loss_diag": 8.23289155960083e-07, "ewc_loss_parallel": 5.103647708892822e-07, "grad_norm": 3.8387882709503174, "learning_rate": 7.54557015684612e-08, "loss": 0.695, "mean_token_accuracy": 0.7937134504318237, "num_tokens": 6861598.0, "step": 179 }, { "epoch": 0.02289785014629182, "ewc_loss": 0.000598907470703125, "ewc_loss_diag": 8.493661880493164e-07, "ewc_loss_parallel": 5.140900611877441e-07, "grad_norm": 3.781656265258789, "learning_rate": 7.587961000423907e-08, "loss": 0.7174, "mean_token_accuracy": 0.7911139130592346, "num_tokens": 6900220.0, "step": 180 }, { "epoch": 0.02302506042488233, "ewc_loss": 0.000606536865234375, "ewc_loss_diag": 8.642673492431641e-07, "ewc_loss_parallel": 5.178153514862061e-07, "grad_norm": 3.9102816581726074, "learning_rate": 7.630351844001694e-08, "loss": 0.8257, "mean_token_accuracy": 0.7598578333854675, "num_tokens": 6946983.0, "step": 181 }, { "epoch": 0.02315227070347284, "ewc_loss": 0.00061798095703125, "ewc_loss_diag": 8.828938007354736e-07, "ewc_loss_parallel": 5.289912223815918e-07, "grad_norm": 3.937265157699585, "learning_rate": 7.672742687579482e-08, "loss": 0.73, "mean_token_accuracy": 0.787688136100769, "num_tokens": 6986019.0, "step": 182 }, { "epoch": 0.02327948098206335, "ewc_loss": 0.000640869140625, "ewc_loss_diag": 9.499490261077881e-07, "ewc_loss_parallel": 5.438923835754395e-07, "grad_norm": 4.074675559997559, "learning_rate": 7.715133531157269e-08, "loss": 0.7042, "mean_token_accuracy": 0.7890062928199768, "num_tokens": 7020722.0, "step": 183 }, { "epoch": 0.02340669126065386, "ewc_loss": 0.00066375732421875, "ewc_loss_diag": 9.685754776000977e-07, "ewc_loss_parallel": 5.699694156646729e-07, "grad_norm": 4.071313381195068, "learning_rate": 7.757524374735056e-08, "loss": 0.7756, "mean_token_accuracy": 0.7740613222122192, "num_tokens": 7062691.0, "step": 184 }, { "epoch": 0.02353390153924437, "ewc_loss": 0.0006866455078125, "ewc_loss_diag": 9.685754776000977e-07, "ewc_loss_parallel": 5.923211574554443e-07, "grad_norm": 3.9322669506073, "learning_rate": 7.799915218312844e-08, "loss": 0.7231, "mean_token_accuracy": 0.7886011600494385, "num_tokens": 7101090.0, "step": 185 }, { "epoch": 0.02366111181783488, "ewc_loss": 0.00070953369140625, "ewc_loss_diag": 9.834766387939453e-07, "ewc_loss_parallel": 6.109476089477539e-07, "grad_norm": 3.8921639919281006, "learning_rate": 7.842306061890631e-08, "loss": 0.6964, "mean_token_accuracy": 0.7910017371177673, "num_tokens": 7139891.0, "step": 186 }, { "epoch": 0.023788322096425393, "ewc_loss": 0.000720977783203125, "ewc_loss_diag": 9.909272193908691e-07, "ewc_loss_parallel": 6.221234798431396e-07, "grad_norm": 4.018792152404785, "learning_rate": 7.88469690546842e-08, "loss": 0.7445, "mean_token_accuracy": 0.7820494174957275, "num_tokens": 7179501.0, "step": 187 }, { "epoch": 0.0239155323750159, "ewc_loss": 0.000736236572265625, "ewc_loss_diag": 9.909272193908691e-07, "ewc_loss_parallel": 6.370246410369873e-07, "grad_norm": 3.9722485542297363, "learning_rate": 7.927087749046207e-08, "loss": 0.7137, "mean_token_accuracy": 0.7886053323745728, "num_tokens": 7213809.0, "step": 188 }, { "epoch": 0.02404274265360641, "ewc_loss": 0.00075531005859375, "ewc_loss_diag": 9.611248970031738e-07, "ewc_loss_parallel": 6.593763828277588e-07, "grad_norm": 3.900437593460083, "learning_rate": 7.969478592623994e-08, "loss": 0.7884, "mean_token_accuracy": 0.7667876482009888, "num_tokens": 7254493.0, "step": 189 }, { "epoch": 0.024169952932196922, "ewc_loss": 0.00078582763671875, "ewc_loss_diag": 9.685754776000977e-07, "ewc_loss_parallel": 6.891787052154541e-07, "grad_norm": 4.413786888122559, "learning_rate": 8.011869436201781e-08, "loss": 0.6464, "mean_token_accuracy": 0.8093986511230469, "num_tokens": 7292643.0, "step": 190 }, { "epoch": 0.024297163210787433, "ewc_loss": 0.000827789306640625, "ewc_loss_diag": 9.760260581970215e-07, "ewc_loss_parallel": 7.301568984985352e-07, "grad_norm": 4.331425666809082, "learning_rate": 8.054260279779568e-08, "loss": 0.7692, "mean_token_accuracy": 0.7767730951309204, "num_tokens": 7329810.0, "step": 191 }, { "epoch": 0.02442437348937794, "ewc_loss": 0.00084686279296875, "ewc_loss_diag": 9.834766387939453e-07, "ewc_loss_parallel": 7.487833499908447e-07, "grad_norm": 4.212569713592529, "learning_rate": 8.096651123357356e-08, "loss": 0.7551, "mean_token_accuracy": 0.7781423926353455, "num_tokens": 7367630.0, "step": 192 }, { "epoch": 0.02455158376796845, "ewc_loss": 0.000881195068359375, "ewc_loss_diag": 9.98377799987793e-07, "ewc_loss_parallel": 7.82310962677002e-07, "grad_norm": 3.8636345863342285, "learning_rate": 8.139041966935143e-08, "loss": 0.7365, "mean_token_accuracy": 0.7828733921051025, "num_tokens": 7411580.0, "step": 193 }, { "epoch": 0.024678794046558962, "ewc_loss": 0.00090789794921875, "ewc_loss_diag": 1.0058283805847168e-06, "ewc_loss_parallel": 8.083879947662354e-07, "grad_norm": 4.088955879211426, "learning_rate": 8.181432810512929e-08, "loss": 0.6905, "mean_token_accuracy": 0.7959448099136353, "num_tokens": 7451631.0, "step": 194 }, { "epoch": 0.024806004325149473, "ewc_loss": 0.0009307861328125, "ewc_loss_diag": 1.0132789611816406e-06, "ewc_loss_parallel": 8.307397365570068e-07, "grad_norm": 4.041158676147461, "learning_rate": 8.223823654090716e-08, "loss": 0.6995, "mean_token_accuracy": 0.7924063205718994, "num_tokens": 7493645.0, "step": 195 }, { "epoch": 0.024933214603739984, "ewc_loss": 0.000965118408203125, "ewc_loss_diag": 1.0281801223754883e-06, "ewc_loss_parallel": 8.605420589447021e-07, "grad_norm": 4.548884868621826, "learning_rate": 8.266214497668503e-08, "loss": 0.6598, "mean_token_accuracy": 0.8063635230064392, "num_tokens": 7538042.0, "step": 196 }, { "epoch": 0.02506042488233049, "ewc_loss": 0.001007080078125, "ewc_loss_diag": 1.043081283569336e-06, "ewc_loss_parallel": 9.052455425262451e-07, "grad_norm": 4.205317497253418, "learning_rate": 8.30860534124629e-08, "loss": 0.7227, "mean_token_accuracy": 0.7859359979629517, "num_tokens": 7575374.0, "step": 197 }, { "epoch": 0.025187635160921002, "ewc_loss": 0.00104522705078125, "ewc_loss_diag": 1.0579824447631836e-06, "ewc_loss_parallel": 9.424984455108643e-07, "grad_norm": 4.567849636077881, "learning_rate": 8.350996184824078e-08, "loss": 0.7147, "mean_token_accuracy": 0.7875205278396606, "num_tokens": 7606880.0, "step": 198 }, { "epoch": 0.025314845439511513, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.0728836059570312e-06, "ewc_loss_parallel": 9.760260581970215e-07, "grad_norm": 4.166138648986816, "learning_rate": 8.393387028401865e-08, "loss": 0.6926, "mean_token_accuracy": 0.7965306043624878, "num_tokens": 7644840.0, "step": 199 }, { "epoch": 0.025442055718102024, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 1.0952353477478027e-06, "ewc_loss_parallel": 1.0207295417785645e-06, "grad_norm": 4.3785600662231445, "learning_rate": 8.435777871979652e-08, "loss": 0.7334, "mean_token_accuracy": 0.7823259830474854, "num_tokens": 7683856.0, "step": 200 }, { "epoch": 0.02556926599669253, "ewc_loss": 0.00113677978515625, "ewc_loss_diag": 1.1026859283447266e-06, "ewc_loss_parallel": 1.0281801223754883e-06, "grad_norm": 5.800936222076416, "learning_rate": 8.47816871555744e-08, "loss": 0.6911, "mean_token_accuracy": 0.7959425449371338, "num_tokens": 7715306.0, "step": 201 }, { "epoch": 0.025696476275283042, "ewc_loss": 0.0011749267578125, "ewc_loss_diag": 1.1548399925231934e-06, "ewc_loss_parallel": 1.0579824447631836e-06, "grad_norm": 4.535664081573486, "learning_rate": 8.520559559135227e-08, "loss": 0.6514, "mean_token_accuracy": 0.8059720396995544, "num_tokens": 7752442.0, "step": 202 }, { "epoch": 0.025823686553873553, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 1.1846423149108887e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 4.390732765197754, "learning_rate": 8.562950402713014e-08, "loss": 0.7354, "mean_token_accuracy": 0.7861988544464111, "num_tokens": 7799891.0, "step": 203 }, { "epoch": 0.025950896832464064, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 1.214444637298584e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 4.427560806274414, "learning_rate": 8.605341246290801e-08, "loss": 0.724, "mean_token_accuracy": 0.7829304337501526, "num_tokens": 7830696.0, "step": 204 }, { "epoch": 0.026078107111054575, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 1.2218952178955078e-06, "ewc_loss_parallel": 1.087784767150879e-06, "grad_norm": 4.163580417633057, "learning_rate": 8.647732089868589e-08, "loss": 0.6945, "mean_token_accuracy": 0.7961305975914001, "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": 3.8944339752197266, "learning_rate": 8.690122933446376e-08, "loss": 0.7182, "mean_token_accuracy": 0.7885095477104187, "num_tokens": 7914418.0, "step": 206 }, { "epoch": 0.026332527668235593, "ewc_loss": 0.001190185546875, "ewc_loss_diag": 1.259148120880127e-06, "ewc_loss_parallel": 1.0654330253601074e-06, "grad_norm": 3.8908472061157227, "learning_rate": 8.732513777024163e-08, "loss": 0.656, "mean_token_accuracy": 0.8097414970397949, "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": 4.131453037261963, "learning_rate": 8.77490462060195e-08, "loss": 0.7049, "mean_token_accuracy": 0.7945982217788696, "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": 6.629356861114502, "learning_rate": 8.817295464179738e-08, "loss": 0.6685, "mean_token_accuracy": 0.7977147102355957, "num_tokens": 8033314.0, "step": 209 }, { "epoch": 0.026714158504007122, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 1.296401023864746e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 4.1345415115356445, "learning_rate": 8.859686307757525e-08, "loss": 0.6872, "mean_token_accuracy": 0.7967211008071899, "num_tokens": 8069362.0, "step": 210 }, { "epoch": 0.026841368782597633, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 1.30385160446167e-06, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 4.275933265686035, "learning_rate": 8.902077151335312e-08, "loss": 0.7198, "mean_token_accuracy": 0.7847492098808289, "num_tokens": 8109126.0, "step": 211 }, { "epoch": 0.026968579061188144, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 1.3932585716247559e-06, "ewc_loss_parallel": 1.1548399925231934e-06, "grad_norm": 4.299740791320801, "learning_rate": 8.944467994913098e-08, "loss": 0.7732, "mean_token_accuracy": 0.7765178680419922, "num_tokens": 8143083.0, "step": 212 }, { "epoch": 0.027095789339778655, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 1.4156103134155273e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 4.8568902015686035, "learning_rate": 8.986858838490885e-08, "loss": 0.6232, "mean_token_accuracy": 0.808350682258606, "num_tokens": 8176636.0, "step": 213 }, { "epoch": 0.027222999618369163, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 1.4230608940124512e-06, "ewc_loss_parallel": 1.1548399925231934e-06, "grad_norm": 4.352575302124023, "learning_rate": 9.029249682068673e-08, "loss": 0.683, "mean_token_accuracy": 0.7961649298667908, "num_tokens": 8220262.0, "step": 214 }, { "epoch": 0.027350209896959674, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 1.3783574104309082e-06, "ewc_loss_parallel": 1.1548399925231934e-06, "grad_norm": 4.186960220336914, "learning_rate": 9.07164052564646e-08, "loss": 0.7009, "mean_token_accuracy": 0.7928199172019958, "num_tokens": 8256303.0, "step": 215 }, { "epoch": 0.027477420175550184, "ewc_loss": 0.00127410888671875, "ewc_loss_diag": 1.385807991027832e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 4.441693305969238, "learning_rate": 9.114031369224247e-08, "loss": 0.6666, "mean_token_accuracy": 0.8038084506988525, "num_tokens": 8293888.0, "step": 216 }, { "epoch": 0.027604630454140695, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 1.4007091522216797e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 4.814620494842529, "learning_rate": 9.156422212802034e-08, "loss": 0.6282, "mean_token_accuracy": 0.8115034103393555, "num_tokens": 8335368.0, "step": 217 }, { "epoch": 0.027731840732731206, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 1.4156103134155273e-06, "ewc_loss_parallel": 1.1548399925231934e-06, "grad_norm": 6.534596920013428, "learning_rate": 9.198813056379822e-08, "loss": 0.6835, "mean_token_accuracy": 0.7985401749610901, "num_tokens": 8370338.0, "step": 218 }, { "epoch": 0.027859051011321714, "ewc_loss": 0.00136566162109375, "ewc_loss_diag": 1.4230608940124512e-06, "ewc_loss_parallel": 1.2218952178955078e-06, "grad_norm": 5.172207832336426, "learning_rate": 9.241203899957609e-08, "loss": 0.6682, "mean_token_accuracy": 0.7997759580612183, "num_tokens": 8409904.0, "step": 219 }, { "epoch": 0.027986261289912225, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 1.4454126358032227e-06, "ewc_loss_parallel": 1.2293457984924316e-06, "grad_norm": 6.1965203285217285, "learning_rate": 9.283594743535396e-08, "loss": 0.7958, "mean_token_accuracy": 0.7665040493011475, "num_tokens": 8446579.0, "step": 220 }, { "epoch": 0.028113471568502735, "ewc_loss": 0.0013885498046875, "ewc_loss_diag": 1.4603137969970703e-06, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 4.873960971832275, "learning_rate": 9.325985587113183e-08, "loss": 0.643, "mean_token_accuracy": 0.809220552444458, "num_tokens": 8483663.0, "step": 221 }, { "epoch": 0.028240681847093246, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 1.5050172805786133e-06, "ewc_loss_parallel": 1.2069940567016602e-06, "grad_norm": 4.934769630432129, "learning_rate": 9.368376430690971e-08, "loss": 0.6732, "mean_token_accuracy": 0.7986423373222351, "num_tokens": 8516346.0, "step": 222 }, { "epoch": 0.028367892125683754, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 1.5050172805786133e-06, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 4.677161693572998, "learning_rate": 9.410767274268758e-08, "loss": 0.6694, "mean_token_accuracy": 0.8017340302467346, "num_tokens": 8556914.0, "step": 223 }, { "epoch": 0.028495102404274265, "ewc_loss": 0.0012664794921875, "ewc_loss_diag": 1.4901161193847656e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 5.057586193084717, "learning_rate": 9.453158117846545e-08, "loss": 0.6581, "mean_token_accuracy": 0.8011350631713867, "num_tokens": 8596548.0, "step": 224 }, { "epoch": 0.028622312682864776, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 1.5050172805786133e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 4.843634605407715, "learning_rate": 9.495548961424333e-08, "loss": 0.6883, "mean_token_accuracy": 0.7987228631973267, "num_tokens": 8634366.0, "step": 225 }, { "epoch": 0.028749522961455286, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 1.5124678611755371e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 4.642027854919434, "learning_rate": 9.53793980500212e-08, "loss": 0.628, "mean_token_accuracy": 0.810222864151001, "num_tokens": 8666480.0, "step": 226 }, { "epoch": 0.028876733240045797, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.5348196029663086e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 5.248358249664307, "learning_rate": 9.580330648579907e-08, "loss": 0.778, "mean_token_accuracy": 0.7748986482620239, "num_tokens": 8705880.0, "step": 227 }, { "epoch": 0.029003943518636305, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.5422701835632324e-06, "ewc_loss_parallel": 1.080334186553955e-06, "grad_norm": 5.815578937530518, "learning_rate": 9.622721492157694e-08, "loss": 0.6399, "mean_token_accuracy": 0.8083568811416626, "num_tokens": 8743772.0, "step": 228 }, { "epoch": 0.029131153797226816, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 1.5497207641601562e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 5.676499843597412, "learning_rate": 9.665112335735482e-08, "loss": 0.6649, "mean_token_accuracy": 0.7971588373184204, "num_tokens": 8775919.0, "step": 229 }, { "epoch": 0.029258364075817327, "ewc_loss": 0.0012664794921875, "ewc_loss_diag": 1.564621925354004e-06, "ewc_loss_parallel": 1.1175870895385742e-06, "grad_norm": 4.4870100021362305, "learning_rate": 9.707503179313267e-08, "loss": 0.6125, "mean_token_accuracy": 0.8139495849609375, "num_tokens": 8809336.0, "step": 230 }, { "epoch": 0.029385574354407838, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.5944242477416992e-06, "ewc_loss_parallel": 1.0579824447631836e-06, "grad_norm": 4.228444576263428, "learning_rate": 9.749894022891055e-08, "loss": 0.6657, "mean_token_accuracy": 0.7998247146606445, "num_tokens": 8845952.0, "step": 231 }, { "epoch": 0.029512784632998345, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 1.6167759895324707e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 5.201577186584473, "learning_rate": 9.792284866468842e-08, "loss": 0.6859, "mean_token_accuracy": 0.79204261302948, "num_tokens": 8889801.0, "step": 232 }, { "epoch": 0.029639994911588856, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 1.6316771507263184e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 6.061028480529785, "learning_rate": 9.834675710046629e-08, "loss": 0.6453, "mean_token_accuracy": 0.8033465147018433, "num_tokens": 8925429.0, "step": 233 }, { "epoch": 0.029767205190179367, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 1.6316771507263184e-06, "ewc_loss_parallel": 1.043081283569336e-06, "grad_norm": 5.506192207336426, "learning_rate": 9.877066553624416e-08, "loss": 0.6779, "mean_token_accuracy": 0.7959789633750916, "num_tokens": 8963360.0, "step": 234 }, { "epoch": 0.029894415468769878, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 1.6391277313232422e-06, "ewc_loss_parallel": 1.0654330253601074e-06, "grad_norm": 5.50456428527832, "learning_rate": 9.919457397202204e-08, "loss": 0.6887, "mean_token_accuracy": 0.796271800994873, "num_tokens": 8998314.0, "step": 235 }, { "epoch": 0.030021625747360385, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.6987323760986328e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 5.980939865112305, "learning_rate": 9.961848240779991e-08, "loss": 0.7592, "mean_token_accuracy": 0.7766317129135132, "num_tokens": 9034804.0, "step": 236 }, { "epoch": 0.030148836025950896, "ewc_loss": 0.001220703125, "ewc_loss_diag": 1.7285346984863281e-06, "ewc_loss_parallel": 1.043081283569336e-06, "grad_norm": 5.295451641082764, "learning_rate": 1.0004239084357778e-07, "loss": 0.6814, "mean_token_accuracy": 0.7973777055740356, "num_tokens": 9070840.0, "step": 237 }, { "epoch": 0.030276046304541407, "ewc_loss": 0.001190185546875, "ewc_loss_diag": 1.7434358596801758e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 5.598268985748291, "learning_rate": 1.0046629927935566e-07, "loss": 0.6635, "mean_token_accuracy": 0.803107500076294, "num_tokens": 9109370.0, "step": 238 }, { "epoch": 0.030403256583131918, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 1.691281795501709e-06, "ewc_loss_parallel": 9.98377799987793e-07, "grad_norm": 4.305200099945068, "learning_rate": 1.0089020771513353e-07, "loss": 0.6061, "mean_token_accuracy": 0.8188248872756958, "num_tokens": 9148191.0, "step": 239 }, { "epoch": 0.03053046686172243, "ewc_loss": 0.00113677978515625, "ewc_loss_diag": 1.691281795501709e-06, "ewc_loss_parallel": 9.685754776000977e-07, "grad_norm": 4.39298677444458, "learning_rate": 1.013141161509114e-07, "loss": 0.6733, "mean_token_accuracy": 0.7993788719177246, "num_tokens": 9182923.0, "step": 240 }, { "epoch": 0.030657677140312936, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.691281795501709e-06, "ewc_loss_parallel": 9.275972843170166e-07, "grad_norm": 4.501692295074463, "learning_rate": 1.0173802458668927e-07, "loss": 0.6726, "mean_token_accuracy": 0.7958387136459351, "num_tokens": 9215882.0, "step": 241 }, { "epoch": 0.030784887418903447, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.6987323760986328e-06, "ewc_loss_parallel": 9.313225746154785e-07, "grad_norm": 6.786174774169922, "learning_rate": 1.0216193302246715e-07, "loss": 0.705, "mean_token_accuracy": 0.7842627763748169, "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": 4.43392276763916, "learning_rate": 1.0258584145824502e-07, "loss": 0.6266, "mean_token_accuracy": 0.8121048212051392, "num_tokens": 9291146.0, "step": 243 }, { "epoch": 0.03103930797608447, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 1.7210841178894043e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.234943866729736, "learning_rate": 1.0300974989402289e-07, "loss": 0.6189, "mean_token_accuracy": 0.8113886713981628, "num_tokens": 9326403.0, "step": 244 }, { "epoch": 0.031166518254674976, "ewc_loss": 0.0011444091796875, "ewc_loss_diag": 1.7210841178894043e-06, "ewc_loss_parallel": 9.760260581970215e-07, "grad_norm": 4.667349338531494, "learning_rate": 1.0343365832980076e-07, "loss": 0.6717, "mean_token_accuracy": 0.7993391752243042, "num_tokens": 9368491.0, "step": 245 }, { "epoch": 0.03129372853326549, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 1.7285346984863281e-06, "ewc_loss_parallel": 9.424984455108643e-07, "grad_norm": 4.4032816886901855, "learning_rate": 1.0385756676557864e-07, "loss": 0.6675, "mean_token_accuracy": 0.8011418581008911, "num_tokens": 9409041.0, "step": 246 }, { "epoch": 0.031420938811855995, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.735985279083252e-06, "ewc_loss_parallel": 9.126961231231689e-07, "grad_norm": 5.75255012512207, "learning_rate": 1.0428147520135651e-07, "loss": 0.6404, "mean_token_accuracy": 0.8073856234550476, "num_tokens": 9448591.0, "step": 247 }, { "epoch": 0.03154814909044651, "ewc_loss": 0.00112152099609375, "ewc_loss_diag": 1.735985279083252e-06, "ewc_loss_parallel": 9.499490261077881e-07, "grad_norm": 4.671078681945801, "learning_rate": 1.0470538363713437e-07, "loss": 0.6707, "mean_token_accuracy": 0.7977492213249207, "num_tokens": 9486604.0, "step": 248 }, { "epoch": 0.031675359369037016, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 1.7508864402770996e-06, "ewc_loss_parallel": 9.350478649139404e-07, "grad_norm": 4.474292278289795, "learning_rate": 1.0512929207291224e-07, "loss": 0.6765, "mean_token_accuracy": 0.7927331924438477, "num_tokens": 9525285.0, "step": 249 }, { "epoch": 0.03180256964762753, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 1.7583370208740234e-06, "ewc_loss_parallel": 9.164214134216309e-07, "grad_norm": 7.458822250366211, "learning_rate": 1.0555320050869011e-07, "loss": 0.657, "mean_token_accuracy": 0.7980604767799377, "num_tokens": 9557898.0, "step": 250 }, { "epoch": 0.03192977992621804, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 1.7583370208740234e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.134983062744141, "learning_rate": 1.0597710894446799e-07, "loss": 0.7166, "mean_token_accuracy": 0.7818976044654846, "num_tokens": 9590438.0, "step": 251 }, { "epoch": 0.032056990204808546, "ewc_loss": 0.0011444091796875, "ewc_loss_diag": 1.7583370208740234e-06, "ewc_loss_parallel": 9.685754776000977e-07, "grad_norm": 5.7193145751953125, "learning_rate": 1.0640101738024586e-07, "loss": 0.6269, "mean_token_accuracy": 0.8107970952987671, "num_tokens": 9631402.0, "step": 252 }, { "epoch": 0.03218420048339906, "ewc_loss": 0.00113677978515625, "ewc_loss_diag": 1.773238182067871e-06, "ewc_loss_parallel": 9.611248970031738e-07, "grad_norm": 6.901003360748291, "learning_rate": 1.0682492581602373e-07, "loss": 0.6081, "mean_token_accuracy": 0.8168078660964966, "num_tokens": 9663702.0, "step": 253 }, { "epoch": 0.03231141076198957, "ewc_loss": 0.00115203857421875, "ewc_loss_diag": 1.780688762664795e-06, "ewc_loss_parallel": 9.760260581970215e-07, "grad_norm": 5.204631805419922, "learning_rate": 1.072488342518016e-07, "loss": 0.5959, "mean_token_accuracy": 0.8168959021568298, "num_tokens": 9698969.0, "step": 254 }, { "epoch": 0.03243862104058008, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 1.780688762664795e-06, "ewc_loss_parallel": 9.387731552124023e-07, "grad_norm": 5.3088250160217285, "learning_rate": 1.0767274268757948e-07, "loss": 0.6655, "mean_token_accuracy": 0.7977379560470581, "num_tokens": 9738075.0, "step": 255 }, { "epoch": 0.03256583131917059, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.7955899238586426e-06, "ewc_loss_parallel": 9.015202522277832e-07, "grad_norm": 6.001148223876953, "learning_rate": 1.0809665112335735e-07, "loss": 0.6709, "mean_token_accuracy": 0.7999792098999023, "num_tokens": 9782675.0, "step": 256 }, { "epoch": 0.0326930415977611, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 1.8104910850524902e-06, "ewc_loss_parallel": 9.126961231231689e-07, "grad_norm": 5.548971652984619, "learning_rate": 1.0852055955913522e-07, "loss": 0.5705, "mean_token_accuracy": 0.8255511522293091, "num_tokens": 9820585.0, "step": 257 }, { "epoch": 0.03282025187635161, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.8253922462463379e-06, "ewc_loss_parallel": 9.052455425262451e-07, "grad_norm": 5.1674909591674805, "learning_rate": 1.089444679949131e-07, "loss": 0.6948, "mean_token_accuracy": 0.7917953729629517, "num_tokens": 9860693.0, "step": 258 }, { "epoch": 0.03294746215494212, "ewc_loss": 0.00106048583984375, "ewc_loss_diag": 1.8849968910217285e-06, "ewc_loss_parallel": 8.754432201385498e-07, "grad_norm": 4.777732849121094, "learning_rate": 1.0936837643069097e-07, "loss": 0.6284, "mean_token_accuracy": 0.8104645609855652, "num_tokens": 9902629.0, "step": 259 }, { "epoch": 0.03307467243353263, "ewc_loss": 0.00102996826171875, "ewc_loss_diag": 1.8402934074401855e-06, "ewc_loss_parallel": 8.456408977508545e-07, "grad_norm": 4.243898391723633, "learning_rate": 1.0979228486646884e-07, "loss": 0.6002, "mean_token_accuracy": 0.8146100640296936, "num_tokens": 9937304.0, "step": 260 }, { "epoch": 0.03320188271212314, "ewc_loss": 0.00099945068359375, "ewc_loss_diag": 1.862645149230957e-06, "ewc_loss_parallel": 8.158385753631592e-07, "grad_norm": 8.222996711730957, "learning_rate": 1.1021619330224671e-07, "loss": 0.734, "mean_token_accuracy": 0.7805131077766418, "num_tokens": 9976911.0, "step": 261 }, { "epoch": 0.03332909299071365, "ewc_loss": 0.00109100341796875, "ewc_loss_diag": 1.862645149230957e-06, "ewc_loss_parallel": 9.08970832824707e-07, "grad_norm": 4.88337516784668, "learning_rate": 1.1064010173802458e-07, "loss": 0.668, "mean_token_accuracy": 0.7996360659599304, "num_tokens": 10015740.0, "step": 262 }, { "epoch": 0.03345630326930416, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 1.8551945686340332e-06, "ewc_loss_parallel": 8.940696716308594e-07, "grad_norm": 4.751597881317139, "learning_rate": 1.1106401017380246e-07, "loss": 0.6052, "mean_token_accuracy": 0.8172073364257812, "num_tokens": 10053054.0, "step": 263 }, { "epoch": 0.03358351354789467, "ewc_loss": 0.00103759765625, "ewc_loss_diag": 1.8700957298278809e-06, "ewc_loss_parallel": 8.493661880493164e-07, "grad_norm": 8.341340065002441, "learning_rate": 1.1148791860958033e-07, "loss": 0.7013, "mean_token_accuracy": 0.7858648896217346, "num_tokens": 10082088.0, "step": 264 }, { "epoch": 0.03371072382648518, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 9.126961231231689e-07, "grad_norm": 6.924882411956787, "learning_rate": 1.119118270453582e-07, "loss": 0.675, "mean_token_accuracy": 0.7953323125839233, "num_tokens": 10121122.0, "step": 265 }, { "epoch": 0.03383793410507569, "ewc_loss": 0.00115203857421875, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 9.611248970031738e-07, "grad_norm": 5.854437828063965, "learning_rate": 1.1233573548113607e-07, "loss": 0.6691, "mean_token_accuracy": 0.7991302013397217, "num_tokens": 10161017.0, "step": 266 }, { "epoch": 0.0339651443836662, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 1.8998980522155762e-06, "ewc_loss_parallel": 9.201467037200928e-07, "grad_norm": 4.413058280944824, "learning_rate": 1.1275964391691393e-07, "loss": 0.6354, "mean_token_accuracy": 0.8088134527206421, "num_tokens": 10204913.0, "step": 267 }, { "epoch": 0.03409235466225671, "ewc_loss": 0.001007080078125, "ewc_loss_diag": 1.8998980522155762e-06, "ewc_loss_parallel": 8.195638656616211e-07, "grad_norm": 4.073937892913818, "learning_rate": 1.131835523526918e-07, "loss": 0.6828, "mean_token_accuracy": 0.7925182580947876, "num_tokens": 10245154.0, "step": 268 }, { "epoch": 0.03421956494084722, "ewc_loss": 0.0009765625, "ewc_loss_diag": 1.8998980522155762e-06, "ewc_loss_parallel": 7.897615432739258e-07, "grad_norm": 5.292881965637207, "learning_rate": 1.1360746078846968e-07, "loss": 0.6902, "mean_token_accuracy": 0.7918302416801453, "num_tokens": 10283006.0, "step": 269 }, { "epoch": 0.03434677521943773, "ewc_loss": 0.00102996826171875, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.381903171539307e-07, "grad_norm": 5.710906028747559, "learning_rate": 1.1403136922424755e-07, "loss": 0.6498, "mean_token_accuracy": 0.8016414642333984, "num_tokens": 10316689.0, "step": 270 }, { "epoch": 0.03447398549802824, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.866190910339355e-07, "grad_norm": 5.330142021179199, "learning_rate": 1.1445527766002542e-07, "loss": 0.6441, "mean_token_accuracy": 0.8028326630592346, "num_tokens": 10358145.0, "step": 271 }, { "epoch": 0.03460119577661875, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.828938007354736e-07, "grad_norm": 5.439616680145264, "learning_rate": 1.148791860958033e-07, "loss": 0.7452, "mean_token_accuracy": 0.7773452401161194, "num_tokens": 10393424.0, "step": 272 }, { "epoch": 0.034728406055209264, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.791685104370117e-07, "grad_norm": 4.899023532867432, "learning_rate": 1.1530309453158117e-07, "loss": 0.6593, "mean_token_accuracy": 0.8000502586364746, "num_tokens": 10433560.0, "step": 273 }, { "epoch": 0.03485561633379977, "ewc_loss": 0.00104522705078125, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 8.530914783477783e-07, "grad_norm": 5.587172985076904, "learning_rate": 1.1572700296735904e-07, "loss": 0.6561, "mean_token_accuracy": 0.8006470203399658, "num_tokens": 10472024.0, "step": 274 }, { "epoch": 0.03498282661239028, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.791685104370117e-07, "grad_norm": 5.047087669372559, "learning_rate": 1.1615091140313691e-07, "loss": 0.6946, "mean_token_accuracy": 0.7899900674819946, "num_tokens": 10510874.0, "step": 275 }, { "epoch": 0.03511003689098079, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.791685104370117e-07, "grad_norm": 5.382311820983887, "learning_rate": 1.1657481983891479e-07, "loss": 0.6327, "mean_token_accuracy": 0.8085501194000244, "num_tokens": 10552411.0, "step": 276 }, { "epoch": 0.0352372471695713, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 8.940696716308594e-07, "grad_norm": 7.527322769165039, "learning_rate": 1.1699872827469266e-07, "loss": 0.6491, "mean_token_accuracy": 0.8049514293670654, "num_tokens": 10591537.0, "step": 277 }, { "epoch": 0.03536445744816181, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 1.952052116394043e-06, "ewc_loss_parallel": 9.350478649139404e-07, "grad_norm": 5.223037242889404, "learning_rate": 1.1742263671047053e-07, "loss": 0.6566, "mean_token_accuracy": 0.8022271990776062, "num_tokens": 10632228.0, "step": 278 }, { "epoch": 0.03549166772675232, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 1.9669532775878906e-06, "ewc_loss_parallel": 9.126961231231689e-07, "grad_norm": 5.151507377624512, "learning_rate": 1.178465451462484e-07, "loss": 0.7122, "mean_token_accuracy": 0.7835077047348022, "num_tokens": 10672829.0, "step": 279 }, { "epoch": 0.03561887800534283, "ewc_loss": 0.001068115234375, "ewc_loss_diag": 1.9669532775878906e-06, "ewc_loss_parallel": 8.717179298400879e-07, "grad_norm": 5.708591461181641, "learning_rate": 1.1827045358202628e-07, "loss": 0.6862, "mean_token_accuracy": 0.7983187437057495, "num_tokens": 10706146.0, "step": 280 }, { "epoch": 0.035746088283933344, "ewc_loss": 0.00107574462890625, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 8.791685104370117e-07, "grad_norm": 6.119521617889404, "learning_rate": 1.1869436201780415e-07, "loss": 0.6031, "mean_token_accuracy": 0.8178365230560303, "num_tokens": 10744112.0, "step": 281 }, { "epoch": 0.03587329856252385, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 9.126961231231689e-07, "grad_norm": 5.506007194519043, "learning_rate": 1.1911827045358202e-07, "loss": 0.6335, "mean_token_accuracy": 0.8124217987060547, "num_tokens": 10783726.0, "step": 282 }, { "epoch": 0.03600050884111436, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 2.0116567611694336e-06, "ewc_loss_parallel": 9.052455425262451e-07, "grad_norm": 7.328701496124268, "learning_rate": 1.195421788893599e-07, "loss": 0.6986, "mean_token_accuracy": 0.7919679880142212, "num_tokens": 10824476.0, "step": 283 }, { "epoch": 0.036127719119704874, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 9.5367431640625e-07, "grad_norm": 7.661795616149902, "learning_rate": 1.1996608732513778e-07, "loss": 0.6582, "mean_token_accuracy": 0.8011298775672913, "num_tokens": 10859997.0, "step": 284 }, { "epoch": 0.03625492939829538, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 9.760260581970215e-07, "grad_norm": 5.144801139831543, "learning_rate": 1.2038999576091563e-07, "loss": 0.6285, "mean_token_accuracy": 0.8104010820388794, "num_tokens": 10898800.0, "step": 285 }, { "epoch": 0.036382139676885895, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 2.0563602447509766e-06, "ewc_loss_parallel": 8.977949619293213e-07, "grad_norm": 5.790563583374023, "learning_rate": 1.208139041966935e-07, "loss": 0.594, "mean_token_accuracy": 0.8210173845291138, "num_tokens": 10937462.0, "step": 286 }, { "epoch": 0.0365093499554764, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 2.0563602447509766e-06, "ewc_loss_parallel": 8.977949619293213e-07, "grad_norm": 5.151595115661621, "learning_rate": 1.2123781263247137e-07, "loss": 0.6554, "mean_token_accuracy": 0.7987088561058044, "num_tokens": 10970739.0, "step": 287 }, { "epoch": 0.03663656023406691, "ewc_loss": 0.0010833740234375, "ewc_loss_diag": 2.0712614059448242e-06, "ewc_loss_parallel": 8.791685104370117e-07, "grad_norm": 5.458436489105225, "learning_rate": 1.2166172106824924e-07, "loss": 0.7078, "mean_token_accuracy": 0.7930272817611694, "num_tokens": 11003000.0, "step": 288 }, { "epoch": 0.036763770512657425, "ewc_loss": 0.0010986328125, "ewc_loss_diag": 2.086162567138672e-06, "ewc_loss_parallel": 8.903443813323975e-07, "grad_norm": 5.981619834899902, "learning_rate": 1.2208562950402712e-07, "loss": 0.6126, "mean_token_accuracy": 0.8106939792633057, "num_tokens": 11039665.0, "step": 289 }, { "epoch": 0.03689098079124793, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 2.1010637283325195e-06, "ewc_loss_parallel": 9.201467037200928e-07, "grad_norm": 5.25791072845459, "learning_rate": 1.22509537939805e-07, "loss": 0.6398, "mean_token_accuracy": 0.8010847568511963, "num_tokens": 11078368.0, "step": 290 }, { "epoch": 0.03701819106983844, "ewc_loss": 0.00112152099609375, "ewc_loss_diag": 2.1010637283325195e-06, "ewc_loss_parallel": 9.126961231231689e-07, "grad_norm": 5.171594142913818, "learning_rate": 1.2293344637558286e-07, "loss": 0.6657, "mean_token_accuracy": 0.7995405197143555, "num_tokens": 11122654.0, "step": 291 }, { "epoch": 0.037145401348428954, "ewc_loss": 0.0011138916015625, "ewc_loss_diag": 2.130866050720215e-06, "ewc_loss_parallel": 9.015202522277832e-07, "grad_norm": 5.059526443481445, "learning_rate": 1.2335735481136073e-07, "loss": 0.6202, "mean_token_accuracy": 0.8055612444877625, "num_tokens": 11162622.0, "step": 292 }, { "epoch": 0.03727261162701946, "ewc_loss": 0.001129150390625, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 9.164214134216309e-07, "grad_norm": 5.84531307220459, "learning_rate": 1.237812632471386e-07, "loss": 0.6323, "mean_token_accuracy": 0.8082841634750366, "num_tokens": 11202564.0, "step": 293 }, { "epoch": 0.037399821905609976, "ewc_loss": 0.00115966796875, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 9.424984455108643e-07, "grad_norm": 7.360135555267334, "learning_rate": 1.2420517168291648e-07, "loss": 0.651, "mean_token_accuracy": 0.7989121079444885, "num_tokens": 11243422.0, "step": 294 }, { "epoch": 0.03752703218420048, "ewc_loss": 0.00119781494140625, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.286892414093018, "learning_rate": 1.2462908011869435e-07, "loss": 0.632, "mean_token_accuracy": 0.804772138595581, "num_tokens": 11280867.0, "step": 295 }, { "epoch": 0.03765424246279099, "ewc_loss": 0.001190185546875, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 9.760260581970215e-07, "grad_norm": 5.6958723068237305, "learning_rate": 1.2505298855447223e-07, "loss": 0.5807, "mean_token_accuracy": 0.8214846849441528, "num_tokens": 11318454.0, "step": 296 }, { "epoch": 0.037781452741381505, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 9.5367431640625e-07, "grad_norm": 6.146790504455566, "learning_rate": 1.254768969902501e-07, "loss": 0.7215, "mean_token_accuracy": 0.7834737300872803, "num_tokens": 11354721.0, "step": 297 }, { "epoch": 0.03790866301997201, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 9.685754776000977e-07, "grad_norm": 6.442085266113281, "learning_rate": 1.2590080542602797e-07, "loss": 0.6786, "mean_token_accuracy": 0.7953813076019287, "num_tokens": 11385938.0, "step": 298 }, { "epoch": 0.03803587329856253, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.032339096069336, "learning_rate": 1.2632471386180584e-07, "loss": 0.6588, "mean_token_accuracy": 0.8000848293304443, "num_tokens": 11424911.0, "step": 299 }, { "epoch": 0.038163083577153034, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.611248970031738e-07, "grad_norm": 5.80555534362793, "learning_rate": 1.2674862229758372e-07, "loss": 0.6268, "mean_token_accuracy": 0.8099868297576904, "num_tokens": 11460580.0, "step": 300 }, { "epoch": 0.03829029385574354, "ewc_loss": 0.00119781494140625, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 9.760260581970215e-07, "grad_norm": 5.07037353515625, "learning_rate": 1.271725307333616e-07, "loss": 0.617, "mean_token_accuracy": 0.813681960105896, "num_tokens": 11501031.0, "step": 301 }, { "epoch": 0.038417504134334056, "ewc_loss": 0.00116729736328125, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 9.499490261077881e-07, "grad_norm": 5.3212714195251465, "learning_rate": 1.2759643916913946e-07, "loss": 0.627, "mean_token_accuracy": 0.8098306655883789, "num_tokens": 11544053.0, "step": 302 }, { "epoch": 0.03854471441292456, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 2.2798776626586914e-06, "ewc_loss_parallel": 9.5367431640625e-07, "grad_norm": 7.313253879547119, "learning_rate": 1.2802034760491733e-07, "loss": 0.643, "mean_token_accuracy": 0.8033790588378906, "num_tokens": 11576944.0, "step": 303 }, { "epoch": 0.03867192469151508, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 5.525121688842773, "learning_rate": 1.284442560406952e-07, "loss": 0.6932, "mean_token_accuracy": 0.7886897325515747, "num_tokens": 11619735.0, "step": 304 }, { "epoch": 0.038799134970105585, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 6.782010555267334, "learning_rate": 1.2886816447647308e-07, "loss": 0.7306, "mean_token_accuracy": 0.7768211364746094, "num_tokens": 11655025.0, "step": 305 }, { "epoch": 0.03892634524869609, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 5.490482330322266, "learning_rate": 1.2929207291225095e-07, "loss": 0.6276, "mean_token_accuracy": 0.8088676929473877, "num_tokens": 11691970.0, "step": 306 }, { "epoch": 0.03905355552728661, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 5.565639972686768, "learning_rate": 1.2971598134802882e-07, "loss": 0.6429, "mean_token_accuracy": 0.8029436469078064, "num_tokens": 11729561.0, "step": 307 }, { "epoch": 0.039180765805877114, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 9.909272193908691e-07, "grad_norm": 6.891100883483887, "learning_rate": 1.301398897838067e-07, "loss": 0.6621, "mean_token_accuracy": 0.7970594167709351, "num_tokens": 11765707.0, "step": 308 }, { "epoch": 0.03930797608446762, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 4.997998237609863, "learning_rate": 1.3056379821958457e-07, "loss": 0.6708, "mean_token_accuracy": 0.8014549612998962, "num_tokens": 11806223.0, "step": 309 }, { "epoch": 0.039435186363058136, "ewc_loss": 0.00121307373046875, "ewc_loss_diag": 2.2649765014648438e-06, "ewc_loss_parallel": 9.834766387939453e-07, "grad_norm": 5.116710186004639, "learning_rate": 1.3098770665536244e-07, "loss": 0.6257, "mean_token_accuracy": 0.8106960654258728, "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": 6.608870029449463, "learning_rate": 1.3141161509114031e-07, "loss": 0.6105, "mean_token_accuracy": 0.8084971904754639, "num_tokens": 11877973.0, "step": 311 }, { "epoch": 0.03968960692023916, "ewc_loss": 0.00124359130859375, "ewc_loss_diag": 2.294778823852539e-06, "ewc_loss_parallel": 1.0132789611816406e-06, "grad_norm": 5.01624870300293, "learning_rate": 1.3183552352691819e-07, "loss": 0.617, "mean_token_accuracy": 0.8112330436706543, "num_tokens": 11912093.0, "step": 312 }, { "epoch": 0.039816817198829665, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 2.3096799850463867e-06, "ewc_loss_parallel": 9.98377799987793e-07, "grad_norm": 5.119631767272949, "learning_rate": 1.3225943196269603e-07, "loss": 0.6431, "mean_token_accuracy": 0.8080354928970337, "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": 5.653764724731445, "learning_rate": 1.3268334039847393e-07, "loss": 0.62, "mean_token_accuracy": 0.8091210126876831, "num_tokens": 11990414.0, "step": 314 }, { "epoch": 0.04007123775601069, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 2.3245811462402344e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 5.415872097015381, "learning_rate": 1.3310724883425178e-07, "loss": 0.6874, "mean_token_accuracy": 0.7883592844009399, "num_tokens": 12025925.0, "step": 315 }, { "epoch": 0.040198448034601195, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 2.339482307434082e-06, "ewc_loss_parallel": 1.0207295417785645e-06, "grad_norm": 5.735225200653076, "learning_rate": 1.3353115727002968e-07, "loss": 0.6548, "mean_token_accuracy": 0.8006784915924072, "num_tokens": 12068879.0, "step": 316 }, { "epoch": 0.04032565831319171, "ewc_loss": 0.00125885009765625, "ewc_loss_diag": 2.339482307434082e-06, "ewc_loss_parallel": 1.0207295417785645e-06, "grad_norm": 4.876884460449219, "learning_rate": 1.3395506570580752e-07, "loss": 0.6211, "mean_token_accuracy": 0.809016227722168, "num_tokens": 12111931.0, "step": 317 }, { "epoch": 0.040452868591782216, "ewc_loss": 0.00124359130859375, "ewc_loss_diag": 2.339482307434082e-06, "ewc_loss_parallel": 1.0058283805847168e-06, "grad_norm": 6.288454532623291, "learning_rate": 1.3437897414158542e-07, "loss": 0.6896, "mean_token_accuracy": 0.7876021862030029, "num_tokens": 12140431.0, "step": 318 }, { "epoch": 0.040580078870372724, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 2.339482307434082e-06, "ewc_loss_parallel": 1.0505318641662598e-06, "grad_norm": 5.715543746948242, "learning_rate": 1.3480288257736327e-07, "loss": 0.6185, "mean_token_accuracy": 0.8091955780982971, "num_tokens": 12173932.0, "step": 319 }, { "epoch": 0.04070728914896324, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 2.3543834686279297e-06, "ewc_loss_parallel": 1.0579824447631836e-06, "grad_norm": 4.972932815551758, "learning_rate": 1.3522679101314117e-07, "loss": 0.5963, "mean_token_accuracy": 0.8146546483039856, "num_tokens": 12210040.0, "step": 320 }, { "epoch": 0.040834499427553746, "ewc_loss": 0.00127410888671875, "ewc_loss_diag": 2.384185791015625e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 5.684479713439941, "learning_rate": 1.35650699448919e-07, "loss": 0.6292, "mean_token_accuracy": 0.8084113597869873, "num_tokens": 12248556.0, "step": 321 }, { "epoch": 0.04096170970614425, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.3990869522094727e-06, "ewc_loss_parallel": 1.043081283569336e-06, "grad_norm": 5.119485378265381, "learning_rate": 1.360746078846969e-07, "loss": 0.6181, "mean_token_accuracy": 0.8096388578414917, "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": 4.928602695465088, "learning_rate": 1.3649851632047476e-07, "loss": 0.6714, "mean_token_accuracy": 0.7951151728630066, "num_tokens": 12326964.0, "step": 323 }, { "epoch": 0.041216130263325275, "ewc_loss": 0.00128173828125, "ewc_loss_diag": 2.428889274597168e-06, "ewc_loss_parallel": 1.0356307029724121e-06, "grad_norm": 5.292009353637695, "learning_rate": 1.3692242475625266e-07, "loss": 0.5802, "mean_token_accuracy": 0.8237448334693909, "num_tokens": 12366541.0, "step": 324 }, { "epoch": 0.04134334054191579, "ewc_loss": 0.00130462646484375, "ewc_loss_diag": 2.4437904357910156e-06, "ewc_loss_parallel": 1.0579824447631836e-06, "grad_norm": 5.826111793518066, "learning_rate": 1.373463331920305e-07, "loss": 0.6417, "mean_token_accuracy": 0.8091608285903931, "num_tokens": 12405664.0, "step": 325 }, { "epoch": 0.0414705508205063, "ewc_loss": 0.0013427734375, "ewc_loss_diag": 2.4437904357910156e-06, "ewc_loss_parallel": 1.0952353477478027e-06, "grad_norm": 5.973715782165527, "learning_rate": 1.377702416278084e-07, "loss": 0.7058, "mean_token_accuracy": 0.7862036824226379, "num_tokens": 12445039.0, "step": 326 }, { "epoch": 0.041597761099096804, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 1.1175870895385742e-06, "grad_norm": 5.443739891052246, "learning_rate": 1.3819415006358625e-07, "loss": 0.5761, "mean_token_accuracy": 0.8237067461013794, "num_tokens": 12485481.0, "step": 327 }, { "epoch": 0.04172497137768732, "ewc_loss": 0.00135040283203125, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 5.468579292297363, "learning_rate": 1.3861805849936415e-07, "loss": 0.6012, "mean_token_accuracy": 0.8189699649810791, "num_tokens": 12530272.0, "step": 328 }, { "epoch": 0.041852181656277826, "ewc_loss": 0.0013427734375, "ewc_loss_diag": 2.473592758178711e-06, "ewc_loss_parallel": 1.0952353477478027e-06, "grad_norm": 6.452935695648193, "learning_rate": 1.39041966935142e-07, "loss": 0.6471, "mean_token_accuracy": 0.8021843433380127, "num_tokens": 12565854.0, "step": 329 }, { "epoch": 0.04197939193486834, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.473592758178711e-06, "ewc_loss_parallel": 1.125037670135498e-06, "grad_norm": 5.037304401397705, "learning_rate": 1.394658753709199e-07, "loss": 0.6329, "mean_token_accuracy": 0.8075199127197266, "num_tokens": 12607313.0, "step": 330 }, { "epoch": 0.04210660221345885, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.4884939193725586e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 5.521357536315918, "learning_rate": 1.3988978380669774e-07, "loss": 0.6707, "mean_token_accuracy": 0.7955096364021301, "num_tokens": 12642057.0, "step": 331 }, { "epoch": 0.042233812492049355, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.4884939193725586e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 5.512545585632324, "learning_rate": 1.403136922424756e-07, "loss": 0.6621, "mean_token_accuracy": 0.7980484962463379, "num_tokens": 12684118.0, "step": 332 }, { "epoch": 0.04236102277063987, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.5033950805664062e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 5.717811584472656, "learning_rate": 1.4073760067825348e-07, "loss": 0.6267, "mean_token_accuracy": 0.8078014850616455, "num_tokens": 12721945.0, "step": 333 }, { "epoch": 0.04248823304923038, "ewc_loss": 0.00138092041015625, "ewc_loss_diag": 2.5033950805664062e-06, "ewc_loss_parallel": 1.1324882507324219e-06, "grad_norm": 4.95211935043335, "learning_rate": 1.4116150911403136e-07, "loss": 0.6407, "mean_token_accuracy": 0.8040995597839355, "num_tokens": 12764175.0, "step": 334 }, { "epoch": 0.04261544332782089, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.518296241760254e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 5.089048385620117, "learning_rate": 1.4158541754980923e-07, "loss": 0.5909, "mean_token_accuracy": 0.8206305503845215, "num_tokens": 12801600.0, "step": 335 }, { "epoch": 0.0427426536064114, "ewc_loss": 0.0013580322265625, "ewc_loss_diag": 2.518296241760254e-06, "ewc_loss_parallel": 1.1026859283447266e-06, "grad_norm": 7.792636394500732, "learning_rate": 1.420093259855871e-07, "loss": 0.6235, "mean_token_accuracy": 0.8148441314697266, "num_tokens": 12843584.0, "step": 336 }, { "epoch": 0.042869863885001906, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.5331974029541016e-06, "ewc_loss_parallel": 1.214444637298584e-06, "grad_norm": 6.655272483825684, "learning_rate": 1.4243323442136497e-07, "loss": 0.6372, "mean_token_accuracy": 0.8053706288337708, "num_tokens": 12878105.0, "step": 337 }, { "epoch": 0.04299707416359242, "ewc_loss": 0.0014801025390625, "ewc_loss_diag": 2.5331974029541016e-06, "ewc_loss_parallel": 1.2293457984924316e-06, "grad_norm": 4.739358901977539, "learning_rate": 1.4285714285714285e-07, "loss": 0.6443, "mean_token_accuracy": 0.8037247657775879, "num_tokens": 12911712.0, "step": 338 }, { "epoch": 0.04312428444218293, "ewc_loss": 0.0013885498046875, "ewc_loss_diag": 2.5480985641479492e-06, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 5.145815372467041, "learning_rate": 1.4328105129292072e-07, "loss": 0.6253, "mean_token_accuracy": 0.810230016708374, "num_tokens": 12951303.0, "step": 339 }, { "epoch": 0.043251494720773435, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.562999725341797e-06, "ewc_loss_parallel": 1.1101365089416504e-06, "grad_norm": 6.510986804962158, "learning_rate": 1.437049597286986e-07, "loss": 0.6123, "mean_token_accuracy": 0.8127967119216919, "num_tokens": 12990544.0, "step": 340 }, { "epoch": 0.04337870499936395, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.592802047729492e-06, "ewc_loss_parallel": 1.2069940567016602e-06, "grad_norm": 7.478458404541016, "learning_rate": 1.4412886816447646e-07, "loss": 0.6715, "mean_token_accuracy": 0.7956435680389404, "num_tokens": 13023766.0, "step": 341 }, { "epoch": 0.04350591527795446, "ewc_loss": 0.00152587890625, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 6.06085205078125, "learning_rate": 1.4455277660025434e-07, "loss": 0.5854, "mean_token_accuracy": 0.8206683397293091, "num_tokens": 13056337.0, "step": 342 }, { "epoch": 0.04363312555654497, "ewc_loss": 0.0014801025390625, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 1.2218952178955078e-06, "grad_norm": 5.26509428024292, "learning_rate": 1.449766850360322e-07, "loss": 0.6437, "mean_token_accuracy": 0.80442214012146, "num_tokens": 13097345.0, "step": 343 }, { "epoch": 0.04376033583513548, "ewc_loss": 0.0014190673828125, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 6.443587303161621, "learning_rate": 1.4540059347181008e-07, "loss": 0.6285, "mean_token_accuracy": 0.8079022169113159, "num_tokens": 13127332.0, "step": 344 }, { "epoch": 0.043887546113725986, "ewc_loss": 0.001434326171875, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1771917343139648e-06, "grad_norm": 6.253369331359863, "learning_rate": 1.4582450190758795e-07, "loss": 0.6587, "mean_token_accuracy": 0.797888994216919, "num_tokens": 13170760.0, "step": 345 }, { "epoch": 0.0440147563923165, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.1995434761047363e-06, "grad_norm": 6.376834869384766, "learning_rate": 1.4624841034336583e-07, "loss": 0.6268, "mean_token_accuracy": 0.8105272054672241, "num_tokens": 13207061.0, "step": 346 }, { "epoch": 0.04414196667090701, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 1.1995434761047363e-06, "grad_norm": 5.092131614685059, "learning_rate": 1.466723187791437e-07, "loss": 0.5953, "mean_token_accuracy": 0.8174730539321899, "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": 4.956255912780762, "learning_rate": 1.4709622721492157e-07, "loss": 0.6053, "mean_token_accuracy": 0.8105273842811584, "num_tokens": 13287193.0, "step": 348 }, { "epoch": 0.04439638722808803, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.1175870895385742e-06, "grad_norm": 5.718587875366211, "learning_rate": 1.4752013565069942e-07, "loss": 0.61, "mean_token_accuracy": 0.8129367828369141, "num_tokens": 13323112.0, "step": 349 }, { "epoch": 0.04452359750667854, "ewc_loss": 0.001434326171875, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.1771917343139648e-06, "grad_norm": 5.860605239868164, "learning_rate": 1.4794404408647732e-07, "loss": 0.606, "mean_token_accuracy": 0.8152509331703186, "num_tokens": 13359653.0, "step": 350 }, { "epoch": 0.04465080778526905, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.2069940567016602e-06, "grad_norm": 5.407392501831055, "learning_rate": 1.4836795252225516e-07, "loss": 0.5762, "mean_token_accuracy": 0.8204832673072815, "num_tokens": 13400571.0, "step": 351 }, { "epoch": 0.04477801806385956, "ewc_loss": 0.0014495849609375, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 1.1846423149108887e-06, "grad_norm": 5.36708927154541, "learning_rate": 1.4879186095803306e-07, "loss": 0.5543, "mean_token_accuracy": 0.8241540193557739, "num_tokens": 13437656.0, "step": 352 }, { "epoch": 0.04490522834245007, "ewc_loss": 0.001434326171875, "ewc_loss_diag": 2.652406692504883e-06, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 5.357035160064697, "learning_rate": 1.492157693938109e-07, "loss": 0.613, "mean_token_accuracy": 0.8124445676803589, "num_tokens": 13474688.0, "step": 353 }, { "epoch": 0.04503243862104058, "ewc_loss": 0.00142669677734375, "ewc_loss_diag": 2.652406692504883e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.199581146240234, "learning_rate": 1.496396778295888e-07, "loss": 0.5511, "mean_token_accuracy": 0.829108715057373, "num_tokens": 13518256.0, "step": 354 }, { "epoch": 0.04515964889963109, "ewc_loss": 0.00142669677734375, "ewc_loss_diag": 2.652406692504883e-06, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 5.238071918487549, "learning_rate": 1.5006358626536665e-07, "loss": 0.6377, "mean_token_accuracy": 0.8067688941955566, "num_tokens": 13554315.0, "step": 355 }, { "epoch": 0.0452868591782216, "ewc_loss": 0.00144195556640625, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 1.1771917343139648e-06, "grad_norm": 5.09042501449585, "learning_rate": 1.5048749470114455e-07, "loss": 0.6591, "mean_token_accuracy": 0.7996935248374939, "num_tokens": 13594388.0, "step": 356 }, { "epoch": 0.04541406945681211, "ewc_loss": 0.0014495849609375, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 1.1846423149108887e-06, "grad_norm": 4.999022960662842, "learning_rate": 1.509114031369224e-07, "loss": 0.5847, "mean_token_accuracy": 0.8177425265312195, "num_tokens": 13633704.0, "step": 357 }, { "epoch": 0.04554127973540262, "ewc_loss": 0.0014495849609375, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 1.1846423149108887e-06, "grad_norm": 5.008456707000732, "learning_rate": 1.513353115727003e-07, "loss": 0.6345, "mean_token_accuracy": 0.8016539812088013, "num_tokens": 13671183.0, "step": 358 }, { "epoch": 0.04566849001399313, "ewc_loss": 0.00146484375, "ewc_loss_diag": 2.682209014892578e-06, "ewc_loss_parallel": 1.1995434761047363e-06, "grad_norm": 4.870608806610107, "learning_rate": 1.5175922000847814e-07, "loss": 0.5892, "mean_token_accuracy": 0.819269597530365, "num_tokens": 13709640.0, "step": 359 }, { "epoch": 0.04579570029258364, "ewc_loss": 0.00147247314453125, "ewc_loss_diag": 2.6971101760864258e-06, "ewc_loss_parallel": 1.2069940567016602e-06, "grad_norm": 5.988343238830566, "learning_rate": 1.5218312844425604e-07, "loss": 0.6121, "mean_token_accuracy": 0.8058004975318909, "num_tokens": 13751114.0, "step": 360 }, { "epoch": 0.045922910571174154, "ewc_loss": 0.00152587890625, "ewc_loss_diag": 2.7120113372802734e-06, "ewc_loss_parallel": 1.259148120880127e-06, "grad_norm": 5.218984603881836, "learning_rate": 1.526070368800339e-07, "loss": 0.5425, "mean_token_accuracy": 0.8263530731201172, "num_tokens": 13786448.0, "step": 361 }, { "epoch": 0.04605012084976466, "ewc_loss": 0.00151824951171875, "ewc_loss_diag": 2.726912498474121e-06, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 4.8585615158081055, "learning_rate": 1.530309453158118e-07, "loss": 0.5654, "mean_token_accuracy": 0.8291245698928833, "num_tokens": 13827663.0, "step": 362 }, { "epoch": 0.04617733112835517, "ewc_loss": 0.00148773193359375, "ewc_loss_diag": 2.726912498474121e-06, "ewc_loss_parallel": 1.214444637298584e-06, "grad_norm": 5.565487384796143, "learning_rate": 1.5345485375158963e-07, "loss": 0.6446, "mean_token_accuracy": 0.8031125068664551, "num_tokens": 13866078.0, "step": 363 }, { "epoch": 0.04630454140694568, "ewc_loss": 0.00152587890625, "ewc_loss_diag": 2.726912498474121e-06, "ewc_loss_parallel": 1.2516975402832031e-06, "grad_norm": 5.136594772338867, "learning_rate": 1.5387876218736753e-07, "loss": 0.5221, "mean_token_accuracy": 0.8362879753112793, "num_tokens": 13909767.0, "step": 364 }, { "epoch": 0.04643175168553619, "ewc_loss": 0.00150299072265625, "ewc_loss_diag": 2.7567148208618164e-06, "ewc_loss_parallel": 1.2293457984924316e-06, "grad_norm": 6.3803534507751465, "learning_rate": 1.5430267062314538e-07, "loss": 0.6305, "mean_token_accuracy": 0.808434247970581, "num_tokens": 13948937.0, "step": 365 }, { "epoch": 0.0465589619641267, "ewc_loss": 0.00157928466796875, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 1.30385160446167e-06, "grad_norm": 6.069995880126953, "learning_rate": 1.5472657905892328e-07, "loss": 0.6412, "mean_token_accuracy": 0.8025595545768738, "num_tokens": 13984039.0, "step": 366 }, { "epoch": 0.04668617224271721, "ewc_loss": 0.00156402587890625, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 1.2889504432678223e-06, "grad_norm": 5.591684818267822, "learning_rate": 1.5515048749470113e-07, "loss": 0.6233, "mean_token_accuracy": 0.8077625036239624, "num_tokens": 14018162.0, "step": 367 }, { "epoch": 0.04681338252130772, "ewc_loss": 0.0015411376953125, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 4.590114116668701, "learning_rate": 1.55574395930479e-07, "loss": 0.6233, "mean_token_accuracy": 0.8081490993499756, "num_tokens": 14056493.0, "step": 368 }, { "epoch": 0.046940592799898234, "ewc_loss": 0.00147247314453125, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 1.1995434761047363e-06, "grad_norm": 5.586653232574463, "learning_rate": 1.5599830436625687e-07, "loss": 0.6166, "mean_token_accuracy": 0.8112010955810547, "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": 5.468503952026367, "learning_rate": 1.5642221280203474e-07, "loss": 0.6352, "mean_token_accuracy": 0.809244692325592, "num_tokens": 14136240.0, "step": 370 }, { "epoch": 0.04719501335707925, "ewc_loss": 0.0015716552734375, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 5.741735458374023, "learning_rate": 1.5684612123781262e-07, "loss": 0.5461, "mean_token_accuracy": 0.8295325040817261, "num_tokens": 14171010.0, "step": 371 }, { "epoch": 0.04732222363566976, "ewc_loss": 0.0015716552734375, "ewc_loss_diag": 2.7865171432495117e-06, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 5.045907020568848, "learning_rate": 1.572700296735905e-07, "loss": 0.6175, "mean_token_accuracy": 0.8158266544342041, "num_tokens": 14206985.0, "step": 372 }, { "epoch": 0.04744943391426027, "ewc_loss": 0.0015411376953125, "ewc_loss_diag": 2.816319465637207e-06, "ewc_loss_parallel": 1.259148120880127e-06, "grad_norm": 6.437637805938721, "learning_rate": 1.576939381093684e-07, "loss": 0.6154, "mean_token_accuracy": 0.8083369135856628, "num_tokens": 14236728.0, "step": 373 }, { "epoch": 0.047576644192850785, "ewc_loss": 0.0016021728515625, "ewc_loss_diag": 2.816319465637207e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 4.943212032318115, "learning_rate": 1.5811784654514623e-07, "loss": 0.5425, "mean_token_accuracy": 0.8325622081756592, "num_tokens": 14270758.0, "step": 374 }, { "epoch": 0.04770385447144129, "ewc_loss": 0.00154876708984375, "ewc_loss_diag": 2.816319465637207e-06, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 4.691620826721191, "learning_rate": 1.5854175498092413e-07, "loss": 0.6597, "mean_token_accuracy": 0.7946861982345581, "num_tokens": 14315002.0, "step": 375 }, { "epoch": 0.0478310647500318, "ewc_loss": 0.00152587890625, "ewc_loss_diag": 2.8312206268310547e-06, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 6.1130571365356445, "learning_rate": 1.5896566341670198e-07, "loss": 0.6306, "mean_token_accuracy": 0.8025030493736267, "num_tokens": 14353385.0, "step": 376 }, { "epoch": 0.047958275028622314, "ewc_loss": 0.0016021728515625, "ewc_loss_diag": 2.86102294921875e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 6.129441738128662, "learning_rate": 1.5938957185247988e-07, "loss": 0.6632, "mean_token_accuracy": 0.7926347255706787, "num_tokens": 14391702.0, "step": 377 }, { "epoch": 0.04808548530721282, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 2.8461217880249023e-06, "ewc_loss_parallel": 1.3485550880432129e-06, "grad_norm": 6.175600051879883, "learning_rate": 1.5981348028825772e-07, "loss": 0.6309, "mean_token_accuracy": 0.8057848215103149, "num_tokens": 14427876.0, "step": 378 }, { "epoch": 0.048212695585803336, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 2.86102294921875e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 6.384801387786865, "learning_rate": 1.6023738872403562e-07, "loss": 0.7176, "mean_token_accuracy": 0.7786440849304199, "num_tokens": 14460349.0, "step": 379 }, { "epoch": 0.048339905864393844, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 2.8759241104125977e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 5.397586345672607, "learning_rate": 1.6066129715981347e-07, "loss": 0.5674, "mean_token_accuracy": 0.8263351917266846, "num_tokens": 14497763.0, "step": 380 }, { "epoch": 0.04846711614298435, "ewc_loss": 0.00157928466796875, "ewc_loss_diag": 2.905726432800293e-06, "ewc_loss_parallel": 1.2889504432678223e-06, "grad_norm": 4.646280288696289, "learning_rate": 1.6108520559559137e-07, "loss": 0.5611, "mean_token_accuracy": 0.8244212865829468, "num_tokens": 14537172.0, "step": 381 }, { "epoch": 0.048594326421574865, "ewc_loss": 0.00153350830078125, "ewc_loss_diag": 2.905726432800293e-06, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 5.926028251647949, "learning_rate": 1.6150911403136921e-07, "loss": 0.6479, "mean_token_accuracy": 0.8059322237968445, "num_tokens": 14565515.0, "step": 382 }, { "epoch": 0.04872153670016537, "ewc_loss": 0.00160980224609375, "ewc_loss_diag": 2.9206275939941406e-06, "ewc_loss_parallel": 1.3187527656555176e-06, "grad_norm": 4.509210109710693, "learning_rate": 1.619330224671471e-07, "loss": 0.6386, "mean_token_accuracy": 0.804876446723938, "num_tokens": 14608104.0, "step": 383 }, { "epoch": 0.04884874697875588, "ewc_loss": 0.001556396484375, "ewc_loss_diag": 2.905726432800293e-06, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 5.6238112449646, "learning_rate": 1.6235693090292496e-07, "loss": 0.6462, "mean_token_accuracy": 0.8060306310653687, "num_tokens": 14645328.0, "step": 384 }, { "epoch": 0.048975957257346395, "ewc_loss": 0.00159454345703125, "ewc_loss_diag": 2.905726432800293e-06, "ewc_loss_parallel": 1.30385160446167e-06, "grad_norm": 5.767752647399902, "learning_rate": 1.6278083933870286e-07, "loss": 0.6804, "mean_token_accuracy": 0.7878654599189758, "num_tokens": 14678792.0, "step": 385 }, { "epoch": 0.0491031675359369, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 2.9206275939941406e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 4.926165580749512, "learning_rate": 1.632047477744807e-07, "loss": 0.6033, "mean_token_accuracy": 0.8150702714920044, "num_tokens": 14715095.0, "step": 386 }, { "epoch": 0.049230377814527417, "ewc_loss": 0.00160980224609375, "ewc_loss_diag": 2.9653310775756836e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 5.229069232940674, "learning_rate": 1.6362865621025858e-07, "loss": 0.6229, "mean_token_accuracy": 0.8090280294418335, "num_tokens": 14753641.0, "step": 387 }, { "epoch": 0.049357588093117924, "ewc_loss": 0.00160980224609375, "ewc_loss_diag": 2.9653310775756836e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 5.479138374328613, "learning_rate": 1.6405256464603645e-07, "loss": 0.5941, "mean_token_accuracy": 0.8153109550476074, "num_tokens": 14788816.0, "step": 388 }, { "epoch": 0.04948479837170843, "ewc_loss": 0.00164031982421875, "ewc_loss_diag": 2.9802322387695312e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 5.175863265991211, "learning_rate": 1.6447647308181432e-07, "loss": 0.638, "mean_token_accuracy": 0.8029133081436157, "num_tokens": 14821607.0, "step": 389 }, { "epoch": 0.049612008650298946, "ewc_loss": 0.00164031982421875, "ewc_loss_diag": 2.995133399963379e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 4.879348278045654, "learning_rate": 1.649003815175922e-07, "loss": 0.6557, "mean_token_accuracy": 0.8007239699363708, "num_tokens": 14861298.0, "step": 390 }, { "epoch": 0.04973921892888945, "ewc_loss": 0.00160980224609375, "ewc_loss_diag": 2.995133399963379e-06, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 4.9940996170043945, "learning_rate": 1.6532428995337007e-07, "loss": 0.5723, "mean_token_accuracy": 0.8221819996833801, "num_tokens": 14899951.0, "step": 391 }, { "epoch": 0.04986642920747997, "ewc_loss": 0.0016326904296875, "ewc_loss_diag": 3.0100345611572266e-06, "ewc_loss_parallel": 1.3336539268493652e-06, "grad_norm": 5.061216354370117, "learning_rate": 1.6574819838914794e-07, "loss": 0.5528, "mean_token_accuracy": 0.8303505182266235, "num_tokens": 14937169.0, "step": 392 }, { "epoch": 0.049993639486070475, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.0249357223510742e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 5.108641624450684, "learning_rate": 1.661721068249258e-07, "loss": 0.606, "mean_token_accuracy": 0.8125659227371216, "num_tokens": 14975549.0, "step": 393 }, { "epoch": 0.05012084976466098, "ewc_loss": 0.00165557861328125, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 1.3485550880432129e-06, "grad_norm": 5.05653190612793, "learning_rate": 1.6659601526070368e-07, "loss": 0.6332, "mean_token_accuracy": 0.806896448135376, "num_tokens": 15016630.0, "step": 394 }, { "epoch": 0.0502480600432515, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 4.659932613372803, "learning_rate": 1.6701992369648156e-07, "loss": 0.6146, "mean_token_accuracy": 0.8094637989997864, "num_tokens": 15057822.0, "step": 395 }, { "epoch": 0.050375270321842004, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 5.691555500030518, "learning_rate": 1.6744383213225943e-07, "loss": 0.5881, "mean_token_accuracy": 0.8128376007080078, "num_tokens": 15094385.0, "step": 396 }, { "epoch": 0.05050248060043251, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 1.385807991027832e-06, "grad_norm": 5.621016025543213, "learning_rate": 1.678677405680373e-07, "loss": 0.5475, "mean_token_accuracy": 0.8260993957519531, "num_tokens": 15130969.0, "step": 397 }, { "epoch": 0.050629690879023026, "ewc_loss": 0.0017242431640625, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 1.4156103134155273e-06, "grad_norm": 6.030815124511719, "learning_rate": 1.6829164900381518e-07, "loss": 0.5685, "mean_token_accuracy": 0.8190338015556335, "num_tokens": 15166586.0, "step": 398 }, { "epoch": 0.05075690115761353, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.0547380447387695e-06, "ewc_loss_parallel": 1.4379620552062988e-06, "grad_norm": 4.973052501678467, "learning_rate": 1.6871555743959305e-07, "loss": 0.5588, "mean_token_accuracy": 0.8230794668197632, "num_tokens": 15209603.0, "step": 399 }, { "epoch": 0.05088411143620405, "ewc_loss": 0.0016632080078125, "ewc_loss_diag": 3.084540367126465e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 5.324310302734375, "learning_rate": 1.6913946587537092e-07, "loss": 0.5762, "mean_token_accuracy": 0.81915682554245, "num_tokens": 15251542.0, "step": 400 }, { "epoch": 0.051011321714794555, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 1.3709068298339844e-06, "grad_norm": 5.942715167999268, "learning_rate": 1.695633743111488e-07, "loss": 0.5907, "mean_token_accuracy": 0.8187942504882812, "num_tokens": 15288940.0, "step": 401 }, { "epoch": 0.05113853199338506, "ewc_loss": 0.0017242431640625, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 1.4156103134155273e-06, "grad_norm": 5.664756774902344, "learning_rate": 1.6998728274692667e-07, "loss": 0.5582, "mean_token_accuracy": 0.829511821269989, "num_tokens": 15321965.0, "step": 402 }, { "epoch": 0.05126574227197558, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.4156103134155273e-06, "grad_norm": 5.028862953186035, "learning_rate": 1.7041119118270454e-07, "loss": 0.5658, "mean_token_accuracy": 0.8254351615905762, "num_tokens": 15362895.0, "step": 403 }, { "epoch": 0.051392952550566084, "ewc_loss": 0.0016632080078125, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 6.608892917633057, "learning_rate": 1.7083509961848238e-07, "loss": 0.6245, "mean_token_accuracy": 0.8062281608581543, "num_tokens": 15396675.0, "step": 404 }, { "epoch": 0.0515201628291566, "ewc_loss": 0.001739501953125, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 1.4379620552062988e-06, "grad_norm": 5.527659893035889, "learning_rate": 1.7125900805426028e-07, "loss": 0.6452, "mean_token_accuracy": 0.8022164106369019, "num_tokens": 15443070.0, "step": 405 }, { "epoch": 0.051647373107747106, "ewc_loss": 0.001708984375, "ewc_loss_diag": 3.1441450119018555e-06, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 5.091899871826172, "learning_rate": 1.7168291649003813e-07, "loss": 0.5724, "mean_token_accuracy": 0.82097989320755, "num_tokens": 15476747.0, "step": 406 }, { "epoch": 0.051774583386337614, "ewc_loss": 0.00164794921875, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.3336539268493652e-06, "grad_norm": 5.1620283126831055, "learning_rate": 1.7210682492581603e-07, "loss": 0.598, "mean_token_accuracy": 0.8158571124076843, "num_tokens": 15514866.0, "step": 407 }, { "epoch": 0.05190179366492813, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.129243850708008e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 4.962484836578369, "learning_rate": 1.7253073336159387e-07, "loss": 0.575, "mean_token_accuracy": 0.8192010521888733, "num_tokens": 15554245.0, "step": 408 }, { "epoch": 0.052029003943518635, "ewc_loss": 0.00167083740234375, "ewc_loss_diag": 3.1441450119018555e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 4.860107421875, "learning_rate": 1.7295464179737177e-07, "loss": 0.6036, "mean_token_accuracy": 0.8152899146080017, "num_tokens": 15593958.0, "step": 409 }, { "epoch": 0.05215621422210915, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.1739473342895508e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 5.314870834350586, "learning_rate": 1.7337855023314962e-07, "loss": 0.5487, "mean_token_accuracy": 0.8281793594360352, "num_tokens": 15632567.0, "step": 410 }, { "epoch": 0.05228342450069966, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.1739473342895508e-06, "ewc_loss_parallel": 1.3783574104309082e-06, "grad_norm": 6.069867134094238, "learning_rate": 1.7380245866892752e-07, "loss": 0.5585, "mean_token_accuracy": 0.8247781991958618, "num_tokens": 15668255.0, "step": 411 }, { "epoch": 0.052410634779290165, "ewc_loss": 0.00177001953125, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 1.4528632164001465e-06, "grad_norm": 4.928730010986328, "learning_rate": 1.7422636710470536e-07, "loss": 0.6108, "mean_token_accuracy": 0.8096491694450378, "num_tokens": 15705435.0, "step": 412 }, { "epoch": 0.05253784505788068, "ewc_loss": 0.00168609619140625, "ewc_loss_diag": 3.1888484954833984e-06, "ewc_loss_parallel": 1.3634562492370605e-06, "grad_norm": 5.141322135925293, "learning_rate": 1.7465027554048326e-07, "loss": 0.6314, "mean_token_accuracy": 0.804851770401001, "num_tokens": 15744657.0, "step": 413 }, { "epoch": 0.05266505533647119, "ewc_loss": 0.0016937255859375, "ewc_loss_diag": 3.1888484954833984e-06, "ewc_loss_parallel": 1.3709068298339844e-06, "grad_norm": 5.558440685272217, "learning_rate": 1.750741839762611e-07, "loss": 0.575, "mean_token_accuracy": 0.8229812383651733, "num_tokens": 15777374.0, "step": 414 }, { "epoch": 0.052792265615061694, "ewc_loss": 0.0017242431640625, "ewc_loss_diag": 3.1888484954833984e-06, "ewc_loss_parallel": 1.4007091522216797e-06, "grad_norm": 5.012101650238037, "learning_rate": 1.75498092412039e-07, "loss": 0.5466, "mean_token_accuracy": 0.8298355340957642, "num_tokens": 15814047.0, "step": 415 }, { "epoch": 0.05291947589365221, "ewc_loss": 0.00170135498046875, "ewc_loss_diag": 3.203749656677246e-06, "ewc_loss_parallel": 1.3783574104309082e-06, "grad_norm": 4.674650192260742, "learning_rate": 1.7592200084781686e-07, "loss": 0.6284, "mean_token_accuracy": 0.8048226237297058, "num_tokens": 15849110.0, "step": 416 }, { "epoch": 0.053046686172242716, "ewc_loss": 0.001678466796875, "ewc_loss_diag": 3.248453140258789e-06, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 4.876840114593506, "learning_rate": 1.7634590928359475e-07, "loss": 0.5985, "mean_token_accuracy": 0.8170945048332214, "num_tokens": 15888910.0, "step": 417 }, { "epoch": 0.05317389645083323, "ewc_loss": 0.001708984375, "ewc_loss_diag": 3.2782554626464844e-06, "ewc_loss_parallel": 1.3783574104309082e-06, "grad_norm": 4.8604607582092285, "learning_rate": 1.767698177193726e-07, "loss": 0.5847, "mean_token_accuracy": 0.8179404735565186, "num_tokens": 15925604.0, "step": 418 }, { "epoch": 0.05330110672942374, "ewc_loss": 0.0017242431640625, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 4.775112152099609, "learning_rate": 1.771937261551505e-07, "loss": 0.5587, "mean_token_accuracy": 0.8321088552474976, "num_tokens": 15961401.0, "step": 419 }, { "epoch": 0.053428317008014245, "ewc_loss": 0.00171661376953125, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 1.385807991027832e-06, "grad_norm": 4.941331386566162, "learning_rate": 1.7761763459092835e-07, "loss": 0.5893, "mean_token_accuracy": 0.8176237344741821, "num_tokens": 16001029.0, "step": 420 }, { "epoch": 0.05355552728660476, "ewc_loss": 0.00171661376953125, "ewc_loss_diag": 3.3080577850341797e-06, "ewc_loss_parallel": 1.385807991027832e-06, "grad_norm": 4.994843006134033, "learning_rate": 1.7804154302670624e-07, "loss": 0.5706, "mean_token_accuracy": 0.8233729004859924, "num_tokens": 16036014.0, "step": 421 }, { "epoch": 0.05368273756519527, "ewc_loss": 0.00173187255859375, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 1.4007091522216797e-06, "grad_norm": 4.685097694396973, "learning_rate": 1.784654514624841e-07, "loss": 0.5522, "mean_token_accuracy": 0.8250408172607422, "num_tokens": 16075373.0, "step": 422 }, { "epoch": 0.05380994784378578, "ewc_loss": 0.00171661376953125, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 1.385807991027832e-06, "grad_norm": 4.8255181312561035, "learning_rate": 1.7888935989826196e-07, "loss": 0.5945, "mean_token_accuracy": 0.812226414680481, "num_tokens": 16111352.0, "step": 423 }, { "epoch": 0.05393715812237629, "ewc_loss": 0.00173187255859375, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 1.4007091522216797e-06, "grad_norm": 4.880448818206787, "learning_rate": 1.7931326833403984e-07, "loss": 0.6106, "mean_token_accuracy": 0.8093650937080383, "num_tokens": 16148826.0, "step": 424 }, { "epoch": 0.054064368400966796, "ewc_loss": 0.0017547607421875, "ewc_loss_diag": 3.337860107421875e-06, "ewc_loss_parallel": 1.4156103134155273e-06, "grad_norm": 4.537337779998779, "learning_rate": 1.797371767698177e-07, "loss": 0.5184, "mean_token_accuracy": 0.8386929035186768, "num_tokens": 16193491.0, "step": 425 }, { "epoch": 0.05419157867955731, "ewc_loss": 0.00173187255859375, "ewc_loss_diag": 3.337860107421875e-06, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 5.233978748321533, "learning_rate": 1.8016108520559558e-07, "loss": 0.5679, "mean_token_accuracy": 0.8235297799110413, "num_tokens": 16235595.0, "step": 426 }, { "epoch": 0.05431878895814782, "ewc_loss": 0.00185394287109375, "ewc_loss_diag": 3.337860107421875e-06, "ewc_loss_parallel": 1.519918441772461e-06, "grad_norm": 5.130341529846191, "learning_rate": 1.8058499364137345e-07, "loss": 0.6218, "mean_token_accuracy": 0.8118330240249634, "num_tokens": 16274105.0, "step": 427 }, { "epoch": 0.054445999236738325, "ewc_loss": 0.00183868408203125, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 1.5050172805786133e-06, "grad_norm": 5.293807506561279, "learning_rate": 1.8100890207715133e-07, "loss": 0.6655, "mean_token_accuracy": 0.7916710376739502, "num_tokens": 16315886.0, "step": 428 }, { "epoch": 0.05457320951532884, "ewc_loss": 0.00183868408203125, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 1.5050172805786133e-06, "grad_norm": 4.82192850112915, "learning_rate": 1.814328105129292e-07, "loss": 0.6062, "mean_token_accuracy": 0.8108547329902649, "num_tokens": 16353021.0, "step": 429 }, { "epoch": 0.05470041979391935, "ewc_loss": 0.00177001953125, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 1.4379620552062988e-06, "grad_norm": 5.508661270141602, "learning_rate": 1.8185671894870707e-07, "loss": 0.5811, "mean_token_accuracy": 0.8158049583435059, "num_tokens": 16380887.0, "step": 430 }, { "epoch": 0.05482763007250986, "ewc_loss": 0.00188446044921875, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 1.5497207641601562e-06, "grad_norm": 4.828311443328857, "learning_rate": 1.8228062738448494e-07, "loss": 0.6375, "mean_token_accuracy": 0.8038937449455261, "num_tokens": 16421066.0, "step": 431 }, { "epoch": 0.05495484035110037, "ewc_loss": 0.0018157958984375, "ewc_loss_diag": 3.3676624298095703e-06, "ewc_loss_parallel": 1.4826655387878418e-06, "grad_norm": 5.223995685577393, "learning_rate": 1.8270453582026282e-07, "loss": 0.6059, "mean_token_accuracy": 0.8111706972122192, "num_tokens": 16457197.0, "step": 432 }, { "epoch": 0.055082050629690876, "ewc_loss": 0.00185394287109375, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 1.519918441772461e-06, "grad_norm": 4.957885265350342, "learning_rate": 1.831284442560407e-07, "loss": 0.561, "mean_token_accuracy": 0.823858916759491, "num_tokens": 16497385.0, "step": 433 }, { "epoch": 0.05520926090828139, "ewc_loss": 0.0018463134765625, "ewc_loss_diag": 3.3676624298095703e-06, "ewc_loss_parallel": 1.5124678611755371e-06, "grad_norm": 5.152019023895264, "learning_rate": 1.8355235269181856e-07, "loss": 0.576, "mean_token_accuracy": 0.821895956993103, "num_tokens": 16540099.0, "step": 434 }, { "epoch": 0.0553364711868719, "ewc_loss": 0.001861572265625, "ewc_loss_diag": 3.3676624298095703e-06, "ewc_loss_parallel": 1.5273690223693848e-06, "grad_norm": 5.136638164520264, "learning_rate": 1.8397626112759643e-07, "loss": 0.573, "mean_token_accuracy": 0.8212742209434509, "num_tokens": 16572760.0, "step": 435 }, { "epoch": 0.05546368146546241, "ewc_loss": 0.0018768310546875, "ewc_loss_diag": 3.382563591003418e-06, "ewc_loss_parallel": 1.5422701835632324e-06, "grad_norm": 4.801198482513428, "learning_rate": 1.844001695633743e-07, "loss": 0.5318, "mean_token_accuracy": 0.8322291374206543, "num_tokens": 16608549.0, "step": 436 }, { "epoch": 0.05559089174405292, "ewc_loss": 0.001800537109375, "ewc_loss_diag": 3.382563591003418e-06, "ewc_loss_parallel": 1.4677643775939941e-06, "grad_norm": 5.370595455169678, "learning_rate": 1.8482407799915218e-07, "loss": 0.57, "mean_token_accuracy": 0.8226782083511353, "num_tokens": 16643800.0, "step": 437 }, { "epoch": 0.05571810202264343, "ewc_loss": 0.001922607421875, "ewc_loss_diag": 3.382563591003418e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 5.6339216232299805, "learning_rate": 1.8524798643493005e-07, "loss": 0.6428, "mean_token_accuracy": 0.7981739640235901, "num_tokens": 16677948.0, "step": 438 }, { "epoch": 0.05584531230123394, "ewc_loss": 0.00194549560546875, "ewc_loss_diag": 3.4421682357788086e-06, "ewc_loss_parallel": 1.601874828338623e-06, "grad_norm": 5.298819541931152, "learning_rate": 1.8567189487070792e-07, "loss": 0.6158, "mean_token_accuracy": 0.8103312849998474, "num_tokens": 16716905.0, "step": 439 }, { "epoch": 0.05597252257982445, "ewc_loss": 0.0019073486328125, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 1.564621925354004e-06, "grad_norm": 4.945947170257568, "learning_rate": 1.8609580330648577e-07, "loss": 0.5272, "mean_token_accuracy": 0.8338184952735901, "num_tokens": 16756158.0, "step": 440 }, { "epoch": 0.05609973285841496, "ewc_loss": 0.001861572265625, "ewc_loss_diag": 3.471970558166504e-06, "ewc_loss_parallel": 1.5124678611755371e-06, "grad_norm": 5.348945140838623, "learning_rate": 1.8651971174226367e-07, "loss": 0.6189, "mean_token_accuracy": 0.811187744140625, "num_tokens": 16794261.0, "step": 441 }, { "epoch": 0.05622694313700547, "ewc_loss": 0.00191497802734375, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 4.784924507141113, "learning_rate": 1.8694362017804152e-07, "loss": 0.5757, "mean_token_accuracy": 0.8193914890289307, "num_tokens": 16834953.0, "step": 442 }, { "epoch": 0.05635415341559598, "ewc_loss": 0.0018463134765625, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 1.5050172805786133e-06, "grad_norm": 5.028074264526367, "learning_rate": 1.8736752861381941e-07, "loss": 0.534, "mean_token_accuracy": 0.8341338634490967, "num_tokens": 16873954.0, "step": 443 }, { "epoch": 0.05648136369418649, "ewc_loss": 0.00188446044921875, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 1.5422701835632324e-06, "grad_norm": 5.270779132843018, "learning_rate": 1.8779143704959726e-07, "loss": 0.5573, "mean_token_accuracy": 0.822778582572937, "num_tokens": 16906018.0, "step": 444 }, { "epoch": 0.056608573972777, "ewc_loss": 0.001922607421875, "ewc_loss_diag": 3.471970558166504e-06, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 4.972332954406738, "learning_rate": 1.8821534548537516e-07, "loss": 0.6105, "mean_token_accuracy": 0.8116289377212524, "num_tokens": 16938945.0, "step": 445 }, { "epoch": 0.05673578425136751, "ewc_loss": 0.00191497802734375, "ewc_loss_diag": 3.4868717193603516e-06, "ewc_loss_parallel": 1.564621925354004e-06, "grad_norm": 4.920023441314697, "learning_rate": 1.88639253921153e-07, "loss": 0.5685, "mean_token_accuracy": 0.8249292373657227, "num_tokens": 16982214.0, "step": 446 }, { "epoch": 0.05686299452995802, "ewc_loss": 0.00188446044921875, "ewc_loss_diag": 3.5017728805541992e-06, "ewc_loss_parallel": 1.5348196029663086e-06, "grad_norm": 4.975004196166992, "learning_rate": 1.890631623569309e-07, "loss": 0.5769, "mean_token_accuracy": 0.8199224472045898, "num_tokens": 17024367.0, "step": 447 }, { "epoch": 0.05699020480854853, "ewc_loss": 0.00189971923828125, "ewc_loss_diag": 3.5315752029418945e-06, "ewc_loss_parallel": 1.5497207641601562e-06, "grad_norm": 5.681394100189209, "learning_rate": 1.8948707079270875e-07, "loss": 0.5994, "mean_token_accuracy": 0.8090935349464417, "num_tokens": 17057451.0, "step": 448 }, { "epoch": 0.057117415087139044, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.6167759895324707e-06, "grad_norm": 6.26770544052124, "learning_rate": 1.8991097922848665e-07, "loss": 0.5504, "mean_token_accuracy": 0.8270772695541382, "num_tokens": 17094633.0, "step": 449 }, { "epoch": 0.05724462536572955, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.6540288925170898e-06, "grad_norm": 5.583790302276611, "learning_rate": 1.903348876642645e-07, "loss": 0.6056, "mean_token_accuracy": 0.8086795806884766, "num_tokens": 17135892.0, "step": 450 }, { "epoch": 0.05737183564432006, "ewc_loss": 0.00193023681640625, "ewc_loss_diag": 3.5762786865234375e-06, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 5.336742401123047, "learning_rate": 1.907587961000424e-07, "loss": 0.5481, "mean_token_accuracy": 0.8281912803649902, "num_tokens": 17173881.0, "step": 451 }, { "epoch": 0.05749904592291057, "ewc_loss": 0.0019073486328125, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 1.5497207641601562e-06, "grad_norm": 4.812662124633789, "learning_rate": 1.9118270453582024e-07, "loss": 0.597, "mean_token_accuracy": 0.8153854608535767, "num_tokens": 17215668.0, "step": 452 }, { "epoch": 0.05762625620150108, "ewc_loss": 0.0018768310546875, "ewc_loss_diag": 3.5762786865234375e-06, "ewc_loss_parallel": 1.519918441772461e-06, "grad_norm": 5.214322090148926, "learning_rate": 1.9160661297159814e-07, "loss": 0.6213, "mean_token_accuracy": 0.8100459575653076, "num_tokens": 17254339.0, "step": 453 }, { "epoch": 0.057753466480091595, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.591179847717285e-06, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 4.800755023956299, "learning_rate": 1.9203052140737599e-07, "loss": 0.6059, "mean_token_accuracy": 0.8092271089553833, "num_tokens": 17288858.0, "step": 454 }, { "epoch": 0.0578806767586821, "ewc_loss": 0.00189971923828125, "ewc_loss_diag": 3.5762786865234375e-06, "ewc_loss_parallel": 1.5422701835632324e-06, "grad_norm": 5.192131042480469, "learning_rate": 1.9245442984315389e-07, "loss": 0.5824, "mean_token_accuracy": 0.8210352659225464, "num_tokens": 17330601.0, "step": 455 }, { "epoch": 0.05800788703727261, "ewc_loss": 0.00194549560546875, "ewc_loss_diag": 3.591179847717285e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 4.972655773162842, "learning_rate": 1.9287833827893173e-07, "loss": 0.6013, "mean_token_accuracy": 0.8121185898780823, "num_tokens": 17367797.0, "step": 456 }, { "epoch": 0.058135097315863124, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.591179847717285e-06, "ewc_loss_parallel": 1.5944242477416992e-06, "grad_norm": 4.978267669677734, "learning_rate": 1.9330224671470963e-07, "loss": 0.5873, "mean_token_accuracy": 0.8169502019882202, "num_tokens": 17403206.0, "step": 457 }, { "epoch": 0.05826230759445363, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.606081008911133e-06, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 5.173634052276611, "learning_rate": 1.9372615515048748e-07, "loss": 0.6335, "mean_token_accuracy": 0.8068068027496338, "num_tokens": 17437207.0, "step": 458 }, { "epoch": 0.05838951787304414, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.5762786865234375e-06, "ewc_loss_parallel": 1.601874828338623e-06, "grad_norm": 5.057223796844482, "learning_rate": 1.9415006358626535e-07, "loss": 0.5696, "mean_token_accuracy": 0.8243366479873657, "num_tokens": 17473335.0, "step": 459 }, { "epoch": 0.05851672815163465, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 5.479349136352539, "learning_rate": 1.9457397202204322e-07, "loss": 0.5597, "mean_token_accuracy": 0.8213411569595337, "num_tokens": 17505104.0, "step": 460 }, { "epoch": 0.05864393843022516, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.606081008911133e-06, "ewc_loss_parallel": 1.646578311920166e-06, "grad_norm": 5.288272857666016, "learning_rate": 1.949978804578211e-07, "loss": 0.5434, "mean_token_accuracy": 0.831720232963562, "num_tokens": 17539052.0, "step": 461 }, { "epoch": 0.058771148708815675, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.6242265701293945e-06, "grad_norm": 5.348646640777588, "learning_rate": 1.9542178889359897e-07, "loss": 0.6172, "mean_token_accuracy": 0.8067836761474609, "num_tokens": 17578942.0, "step": 462 }, { "epoch": 0.05889835898740618, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 5.249217510223389, "learning_rate": 1.9584569732937684e-07, "loss": 0.5953, "mean_token_accuracy": 0.8183655142784119, "num_tokens": 17611931.0, "step": 463 }, { "epoch": 0.05902556926599669, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.601874828338623e-06, "grad_norm": 4.95209264755249, "learning_rate": 1.962696057651547e-07, "loss": 0.5005, "mean_token_accuracy": 0.8430982828140259, "num_tokens": 17648764.0, "step": 464 }, { "epoch": 0.059152779544587204, "ewc_loss": 0.001953125, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.5869736671447754e-06, "grad_norm": 5.419329643249512, "learning_rate": 1.9669351420093258e-07, "loss": 0.6112, "mean_token_accuracy": 0.8096645474433899, "num_tokens": 17683099.0, "step": 465 }, { "epoch": 0.05927998982317771, "ewc_loss": 0.0019989013671875, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 1.6316771507263184e-06, "grad_norm": 5.303265571594238, "learning_rate": 1.9711742263671046e-07, "loss": 0.5967, "mean_token_accuracy": 0.8191613554954529, "num_tokens": 17727120.0, "step": 466 }, { "epoch": 0.059407200101768226, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.6507844924926758e-06, "ewc_loss_parallel": 1.6167759895324707e-06, "grad_norm": 6.2629170417785645, "learning_rate": 1.9754133107248833e-07, "loss": 0.5515, "mean_token_accuracy": 0.821671187877655, "num_tokens": 17761676.0, "step": 467 }, { "epoch": 0.059534410380358734, "ewc_loss": 0.0020599365234375, "ewc_loss_diag": 3.6954879760742188e-06, "ewc_loss_parallel": 1.6838312149047852e-06, "grad_norm": 5.09023904800415, "learning_rate": 1.979652395082662e-07, "loss": 0.6187, "mean_token_accuracy": 0.8098807334899902, "num_tokens": 17802311.0, "step": 468 }, { "epoch": 0.05966162065894924, "ewc_loss": 0.0019683837890625, "ewc_loss_diag": 3.7103891372680664e-06, "ewc_loss_parallel": 1.601874828338623e-06, "grad_norm": 4.643792629241943, "learning_rate": 1.9838914794404408e-07, "loss": 0.5099, "mean_token_accuracy": 0.8362995386123657, "num_tokens": 17839048.0, "step": 469 }, { "epoch": 0.059788830937539755, "ewc_loss": 0.0019378662109375, "ewc_loss_diag": 3.6954879760742188e-06, "ewc_loss_parallel": 1.564621925354004e-06, "grad_norm": 5.082488536834717, "learning_rate": 1.9881305637982195e-07, "loss": 0.5818, "mean_token_accuracy": 0.8193782567977905, "num_tokens": 17881510.0, "step": 470 }, { "epoch": 0.05991604121613026, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.6954879760742188e-06, "ewc_loss_parallel": 1.6391277313232422e-06, "grad_norm": 5.512065887451172, "learning_rate": 1.9923696481559982e-07, "loss": 0.6066, "mean_token_accuracy": 0.8135406970977783, "num_tokens": 17915080.0, "step": 471 }, { "epoch": 0.06004325149472077, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.7103891372680664e-06, "ewc_loss_parallel": 1.6540288925170898e-06, "grad_norm": 5.300415515899658, "learning_rate": 1.996608732513777e-07, "loss": 0.6398, "mean_token_accuracy": 0.803540050983429, "num_tokens": 17957972.0, "step": 472 }, { "epoch": 0.060170461773311285, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.7103891372680664e-06, "ewc_loss_parallel": 1.6391277313232422e-06, "grad_norm": 5.058114528656006, "learning_rate": 2.0008478168715557e-07, "loss": 0.5105, "mean_token_accuracy": 0.8362783193588257, "num_tokens": 17998008.0, "step": 473 }, { "epoch": 0.06029767205190179, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.7401914596557617e-06, "ewc_loss_parallel": 1.601874828338623e-06, "grad_norm": 4.987092971801758, "learning_rate": 2.0050869012293344e-07, "loss": 0.5179, "mean_token_accuracy": 0.8335692286491394, "num_tokens": 18032427.0, "step": 474 }, { "epoch": 0.060424882330492306, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 4.883942127227783, "learning_rate": 2.009325985587113e-07, "loss": 0.5485, "mean_token_accuracy": 0.8267040252685547, "num_tokens": 18069848.0, "step": 475 }, { "epoch": 0.060552092609082814, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 5.093684673309326, "learning_rate": 2.0135650699448918e-07, "loss": 0.641, "mean_token_accuracy": 0.8007149696350098, "num_tokens": 18109240.0, "step": 476 }, { "epoch": 0.06067930288767332, "ewc_loss": 0.0019989013671875, "ewc_loss_diag": 3.7401914596557617e-06, "ewc_loss_parallel": 1.6242265701293945e-06, "grad_norm": 4.976500988006592, "learning_rate": 2.0178041543026706e-07, "loss": 0.6112, "mean_token_accuracy": 0.811890721321106, "num_tokens": 18146849.0, "step": 477 }, { "epoch": 0.060806513166263836, "ewc_loss": 0.00201416015625, "ewc_loss_diag": 3.769993782043457e-06, "ewc_loss_parallel": 1.6316771507263184e-06, "grad_norm": 5.144163131713867, "learning_rate": 2.022043238660449e-07, "loss": 0.6418, "mean_token_accuracy": 0.8001256585121155, "num_tokens": 18186872.0, "step": 478 }, { "epoch": 0.06093372344485434, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.769993782043457e-06, "ewc_loss_parallel": 1.646578311920166e-06, "grad_norm": 5.035760402679443, "learning_rate": 2.026282323018228e-07, "loss": 0.6172, "mean_token_accuracy": 0.8109430074691772, "num_tokens": 18222299.0, "step": 479 }, { "epoch": 0.06106093372344486, "ewc_loss": 0.0019989013671875, "ewc_loss_diag": 3.769993782043457e-06, "ewc_loss_parallel": 1.6242265701293945e-06, "grad_norm": 4.5985493659973145, "learning_rate": 2.0305214073760065e-07, "loss": 0.6211, "mean_token_accuracy": 0.8111792802810669, "num_tokens": 18261645.0, "step": 480 }, { "epoch": 0.061188144002035365, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.5944242477416992e-06, "grad_norm": 4.876395225524902, "learning_rate": 2.0347604917337855e-07, "loss": 0.5912, "mean_token_accuracy": 0.8163588047027588, "num_tokens": 18292970.0, "step": 481 }, { "epoch": 0.06131535428062587, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.646578311920166e-06, "grad_norm": 4.77957820892334, "learning_rate": 2.038999576091564e-07, "loss": 0.539, "mean_token_accuracy": 0.8339126110076904, "num_tokens": 18331133.0, "step": 482 }, { "epoch": 0.06144256455921639, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6391277313232422e-06, "grad_norm": 4.778593063354492, "learning_rate": 2.043238660449343e-07, "loss": 0.6251, "mean_token_accuracy": 0.8067128658294678, "num_tokens": 18368905.0, "step": 483 }, { "epoch": 0.061569774837806894, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.7997961044311523e-06, "ewc_loss_parallel": 1.646578311920166e-06, "grad_norm": 5.066132545471191, "learning_rate": 2.0474777448071214e-07, "loss": 0.5499, "mean_token_accuracy": 0.8293529152870178, "num_tokens": 18406362.0, "step": 484 }, { "epoch": 0.0616969851163974, "ewc_loss": 0.002044677734375, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.6614794731140137e-06, "grad_norm": 4.857003688812256, "learning_rate": 2.0517168291649004e-07, "loss": 0.587, "mean_token_accuracy": 0.8179429769515991, "num_tokens": 18439246.0, "step": 485 }, { "epoch": 0.061824195394987916, "ewc_loss": 0.0020294189453125, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.646578311920166e-06, "grad_norm": 5.0305657386779785, "learning_rate": 2.0559559135226788e-07, "loss": 0.5909, "mean_token_accuracy": 0.8141506910324097, "num_tokens": 18473408.0, "step": 486 }, { "epoch": 0.06195140567357842, "ewc_loss": 0.002044677734375, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.6614794731140137e-06, "grad_norm": 4.972643852233887, "learning_rate": 2.0601949978804578e-07, "loss": 0.5446, "mean_token_accuracy": 0.8282204270362854, "num_tokens": 18513246.0, "step": 487 }, { "epoch": 0.06207861595216894, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 1.6838312149047852e-06, "grad_norm": 4.9441680908203125, "learning_rate": 2.0644340822382363e-07, "loss": 0.517, "mean_token_accuracy": 0.8366367220878601, "num_tokens": 18549125.0, "step": 488 }, { "epoch": 0.062205826230759445, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6838312149047852e-06, "grad_norm": 5.014993190765381, "learning_rate": 2.0686731665960153e-07, "loss": 0.5665, "mean_token_accuracy": 0.8281582593917847, "num_tokens": 18585306.0, "step": 489 }, { "epoch": 0.06233303650934995, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.691281795501709e-06, "grad_norm": 4.907532215118408, "learning_rate": 2.0729122509537937e-07, "loss": 0.526, "mean_token_accuracy": 0.8358079195022583, "num_tokens": 18624378.0, "step": 490 }, { "epoch": 0.06246024678794047, "ewc_loss": 0.0020904541015625, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6987323760986328e-06, "grad_norm": 5.356276988983154, "learning_rate": 2.0771513353115727e-07, "loss": 0.57, "mean_token_accuracy": 0.8209542036056519, "num_tokens": 18660814.0, "step": 491 }, { "epoch": 0.06258745706653097, "ewc_loss": 0.0021209716796875, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 4.886650085449219, "learning_rate": 2.0813904196693512e-07, "loss": 0.5387, "mean_token_accuracy": 0.832341730594635, "num_tokens": 18702129.0, "step": 492 }, { "epoch": 0.06271466734512149, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.691281795501709e-06, "grad_norm": 5.94556188583374, "learning_rate": 2.0856295040271302e-07, "loss": 0.6102, "mean_token_accuracy": 0.8087677955627441, "num_tokens": 18737496.0, "step": 493 }, { "epoch": 0.06284187762371199, "ewc_loss": 0.002166748046875, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 1.7881393432617188e-06, "grad_norm": 4.823672294616699, "learning_rate": 2.0898685883849086e-07, "loss": 0.5672, "mean_token_accuracy": 0.8258352279663086, "num_tokens": 18779332.0, "step": 494 }, { "epoch": 0.0629690879023025, "ewc_loss": 0.0020599365234375, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 1.6763806343078613e-06, "grad_norm": 4.646958351135254, "learning_rate": 2.0941076727426874e-07, "loss": 0.5527, "mean_token_accuracy": 0.826831042766571, "num_tokens": 18817976.0, "step": 495 }, { "epoch": 0.06309629818089302, "ewc_loss": 0.0020751953125, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 1.6838312149047852e-06, "grad_norm": 4.556861400604248, "learning_rate": 2.098346757100466e-07, "loss": 0.5436, "mean_token_accuracy": 0.8288231492042542, "num_tokens": 18862860.0, "step": 496 }, { "epoch": 0.06322350845948353, "ewc_loss": 0.0020904541015625, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 1.6987323760986328e-06, "grad_norm": 5.210140228271484, "learning_rate": 2.1025858414582448e-07, "loss": 0.5844, "mean_token_accuracy": 0.8172199726104736, "num_tokens": 18899025.0, "step": 497 }, { "epoch": 0.06335071873807403, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 1.773238182067871e-06, "grad_norm": 5.114514350891113, "learning_rate": 2.1068249258160238e-07, "loss": 0.5506, "mean_token_accuracy": 0.8211950063705444, "num_tokens": 18933118.0, "step": 498 }, { "epoch": 0.06347792901666455, "ewc_loss": 0.002166748046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.773238182067871e-06, "grad_norm": 5.030749320983887, "learning_rate": 2.1110640101738023e-07, "loss": 0.5517, "mean_token_accuracy": 0.8288262486457825, "num_tokens": 18969165.0, "step": 499 }, { "epoch": 0.06360513929525506, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 4.801318168640137, "learning_rate": 2.1153030945315813e-07, "loss": 0.5605, "mean_token_accuracy": 0.8236533999443054, "num_tokens": 19003882.0, "step": 500 }, { "epoch": 0.06373234957384556, "ewc_loss": 0.002105712890625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7136335372924805e-06, "grad_norm": 5.486578941345215, "learning_rate": 2.1195421788893597e-07, "loss": 0.5608, "mean_token_accuracy": 0.8260141611099243, "num_tokens": 19037540.0, "step": 501 }, { "epoch": 0.06385955985243608, "ewc_loss": 0.002197265625, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.8030405044555664e-06, "grad_norm": 5.087930202484131, "learning_rate": 2.1237812632471387e-07, "loss": 0.5711, "mean_token_accuracy": 0.8171701431274414, "num_tokens": 19077270.0, "step": 502 }, { "epoch": 0.06398677013102659, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7657876014709473e-06, "grad_norm": 4.931304931640625, "learning_rate": 2.1280203476049172e-07, "loss": 0.6588, "mean_token_accuracy": 0.7922813296318054, "num_tokens": 19117579.0, "step": 503 }, { "epoch": 0.06411398040961709, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 4.735860824584961, "learning_rate": 2.1322594319626962e-07, "loss": 0.5984, "mean_token_accuracy": 0.8129249811172485, "num_tokens": 19156599.0, "step": 504 }, { "epoch": 0.0642411906882076, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 5.545324802398682, "learning_rate": 2.1364985163204746e-07, "loss": 0.5894, "mean_token_accuracy": 0.8156799674034119, "num_tokens": 19187367.0, "step": 505 }, { "epoch": 0.06436840096679812, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 5.3926472663879395, "learning_rate": 2.1407376006782536e-07, "loss": 0.5249, "mean_token_accuracy": 0.8337884545326233, "num_tokens": 19223916.0, "step": 506 }, { "epoch": 0.06449561124538863, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.8477439880371094e-06, "grad_norm": 4.964799404144287, "learning_rate": 2.144976685036032e-07, "loss": 0.5399, "mean_token_accuracy": 0.8313566446304321, "num_tokens": 19260336.0, "step": 507 }, { "epoch": 0.06462282152397913, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7434358596801758e-06, "grad_norm": 4.85211706161499, "learning_rate": 2.149215769393811e-07, "loss": 0.6019, "mean_token_accuracy": 0.8149017691612244, "num_tokens": 19299895.0, "step": 508 }, { "epoch": 0.06475003180256965, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.735985279083252e-06, "grad_norm": 5.070591926574707, "learning_rate": 2.1534548537515895e-07, "loss": 0.5433, "mean_token_accuracy": 0.8283563852310181, "num_tokens": 19336614.0, "step": 509 }, { "epoch": 0.06487724208116016, "ewc_loss": 0.002166748046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.780688762664795e-06, "grad_norm": 5.0289225578308105, "learning_rate": 2.1576939381093685e-07, "loss": 0.5943, "mean_token_accuracy": 0.8162380456924438, "num_tokens": 19378683.0, "step": 510 }, { "epoch": 0.06500445235975066, "ewc_loss": 0.002166748046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.780688762664795e-06, "grad_norm": 4.962836265563965, "learning_rate": 2.161933022467147e-07, "loss": 0.578, "mean_token_accuracy": 0.8163493275642395, "num_tokens": 19422598.0, "step": 511 }, { "epoch": 0.06513166263834118, "ewc_loss": 0.002166748046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.773238182067871e-06, "grad_norm": 4.634310245513916, "learning_rate": 2.166172106824926e-07, "loss": 0.5682, "mean_token_accuracy": 0.8256514072418213, "num_tokens": 19466972.0, "step": 512 }, { "epoch": 0.06525887291693169, "ewc_loss": 0.0021514892578125, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.7583370208740234e-06, "grad_norm": 5.155607223510742, "learning_rate": 2.1704111911827044e-07, "loss": 0.5592, "mean_token_accuracy": 0.8202366232872009, "num_tokens": 19503809.0, "step": 513 }, { "epoch": 0.0653860831955222, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 1.8551945686340332e-06, "grad_norm": 4.800339221954346, "learning_rate": 2.1746502755404831e-07, "loss": 0.614, "mean_token_accuracy": 0.8144731521606445, "num_tokens": 19545713.0, "step": 514 }, { "epoch": 0.06551329347411271, "ewc_loss": 0.002197265625, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.8030405044555664e-06, "grad_norm": 4.918365955352783, "learning_rate": 2.178889359898262e-07, "loss": 0.6007, "mean_token_accuracy": 0.8157640695571899, "num_tokens": 19582272.0, "step": 515 }, { "epoch": 0.06564050375270322, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.817941665649414e-06, "grad_norm": 4.693600654602051, "learning_rate": 2.1831284442560406e-07, "loss": 0.6165, "mean_token_accuracy": 0.8106766939163208, "num_tokens": 19625574.0, "step": 516 }, { "epoch": 0.06576771403129372, "ewc_loss": 0.0021820068359375, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.7881393432617188e-06, "grad_norm": 4.825271129608154, "learning_rate": 2.1873675286138193e-07, "loss": 0.6038, "mean_token_accuracy": 0.8097472190856934, "num_tokens": 19667791.0, "step": 517 }, { "epoch": 0.06589492430988424, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.8253922462463379e-06, "grad_norm": 4.842251300811768, "learning_rate": 2.191606612971598e-07, "loss": 0.6325, "mean_token_accuracy": 0.804649829864502, "num_tokens": 19708289.0, "step": 518 }, { "epoch": 0.06602213458847475, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.8551945686340332e-06, "grad_norm": 4.845822334289551, "learning_rate": 2.1958456973293768e-07, "loss": 0.6115, "mean_token_accuracy": 0.8052230477333069, "num_tokens": 19747444.0, "step": 519 }, { "epoch": 0.06614934486706527, "ewc_loss": 0.00225830078125, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 1.862645149230957e-06, "grad_norm": 4.927950382232666, "learning_rate": 2.2000847816871555e-07, "loss": 0.4999, "mean_token_accuracy": 0.8383371233940125, "num_tokens": 19787061.0, "step": 520 }, { "epoch": 0.06627655514565577, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 4.676397800445557, "learning_rate": 2.2043238660449342e-07, "loss": 0.5413, "mean_token_accuracy": 0.8310458660125732, "num_tokens": 19826893.0, "step": 521 }, { "epoch": 0.06640376542424628, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 1.8477439880371094e-06, "grad_norm": 4.845484256744385, "learning_rate": 2.208562950402713e-07, "loss": 0.5347, "mean_token_accuracy": 0.8282610177993774, "num_tokens": 19864285.0, "step": 522 }, { "epoch": 0.0665309757028368, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 4.753975868225098, "learning_rate": 2.2128020347604917e-07, "loss": 0.5701, "mean_token_accuracy": 0.8222825527191162, "num_tokens": 19903806.0, "step": 523 }, { "epoch": 0.0666581859814273, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8700957298278809e-06, "grad_norm": 5.079737186431885, "learning_rate": 2.2170411191182704e-07, "loss": 0.5477, "mean_token_accuracy": 0.8259521126747131, "num_tokens": 19950382.0, "step": 524 }, { "epoch": 0.06678539626001781, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.9073486328125e-06, "grad_norm": 4.921200275421143, "learning_rate": 2.221280203476049e-07, "loss": 0.5714, "mean_token_accuracy": 0.8182582855224609, "num_tokens": 19984765.0, "step": 525 }, { "epoch": 0.06691260653860832, "ewc_loss": 0.0022735595703125, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 4.7364606857299805, "learning_rate": 2.2255192878338279e-07, "loss": 0.6112, "mean_token_accuracy": 0.8122455477714539, "num_tokens": 20021081.0, "step": 526 }, { "epoch": 0.06703981681719882, "ewc_loss": 0.00225830078125, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.8551945686340332e-06, "grad_norm": 5.1168718338012695, "learning_rate": 2.2297583721916066e-07, "loss": 0.5838, "mean_token_accuracy": 0.8166123628616333, "num_tokens": 20055860.0, "step": 527 }, { "epoch": 0.06716702709578934, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 1.9222497940063477e-06, "grad_norm": 5.248357772827148, "learning_rate": 2.2339974565493853e-07, "loss": 0.5392, "mean_token_accuracy": 0.8317767977714539, "num_tokens": 20095023.0, "step": 528 }, { "epoch": 0.06729423737437985, "ewc_loss": 0.0023345947265625, "ewc_loss_diag": 4.023313522338867e-06, "ewc_loss_parallel": 1.9371509552001953e-06, "grad_norm": 4.878886699676514, "learning_rate": 2.238236540907164e-07, "loss": 0.5933, "mean_token_accuracy": 0.8154295682907104, "num_tokens": 20134620.0, "step": 529 }, { "epoch": 0.06742144765297035, "ewc_loss": 0.00225830078125, "ewc_loss_diag": 4.0531158447265625e-06, "ewc_loss_parallel": 1.862645149230957e-06, "grad_norm": 5.06915283203125, "learning_rate": 2.2424756252649428e-07, "loss": 0.5643, "mean_token_accuracy": 0.8203924298286438, "num_tokens": 20169690.0, "step": 530 }, { "epoch": 0.06754865793156087, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.8924474716186523e-06, "grad_norm": 5.205932140350342, "learning_rate": 2.2467147096227215e-07, "loss": 0.5564, "mean_token_accuracy": 0.8254414796829224, "num_tokens": 20203609.0, "step": 531 }, { "epoch": 0.06767586821015138, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 4.923989295959473, "learning_rate": 2.2509537939805002e-07, "loss": 0.5631, "mean_token_accuracy": 0.8217965364456177, "num_tokens": 20234616.0, "step": 532 }, { "epoch": 0.0678030784887419, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.9073486328125e-06, "grad_norm": 4.797372817993164, "learning_rate": 2.2551928783382787e-07, "loss": 0.5532, "mean_token_accuracy": 0.8234544992446899, "num_tokens": 20271634.0, "step": 533 }, { "epoch": 0.0679302887673324, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.8998980522155762e-06, "grad_norm": 5.424249172210693, "learning_rate": 2.2594319626960577e-07, "loss": 0.5815, "mean_token_accuracy": 0.821482241153717, "num_tokens": 20309683.0, "step": 534 }, { "epoch": 0.06805749904592291, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.766209125518799, "learning_rate": 2.263671047053836e-07, "loss": 0.4952, "mean_token_accuracy": 0.8384103775024414, "num_tokens": 20346013.0, "step": 535 }, { "epoch": 0.06818470932451343, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 1.8998980522155762e-06, "grad_norm": 4.970055103302002, "learning_rate": 2.267910131411615e-07, "loss": 0.553, "mean_token_accuracy": 0.8285727500915527, "num_tokens": 20387613.0, "step": 536 }, { "epoch": 0.06831191960310393, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 1.9073486328125e-06, "grad_norm": 5.15675687789917, "learning_rate": 2.2721492157693936e-07, "loss": 0.5518, "mean_token_accuracy": 0.8266997933387756, "num_tokens": 20417579.0, "step": 537 }, { "epoch": 0.06843912988169444, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 4.713381290435791, "learning_rate": 2.2763883001271726e-07, "loss": 0.5244, "mean_token_accuracy": 0.8328999280929565, "num_tokens": 20455100.0, "step": 538 }, { "epoch": 0.06856634016028496, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.8998980522155762e-06, "grad_norm": 4.657907485961914, "learning_rate": 2.280627384484951e-07, "loss": 0.5864, "mean_token_accuracy": 0.8142186403274536, "num_tokens": 20496458.0, "step": 539 }, { "epoch": 0.06869355043887546, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 1.8924474716186523e-06, "grad_norm": 5.0407867431640625, "learning_rate": 2.28486646884273e-07, "loss": 0.5605, "mean_token_accuracy": 0.8222222328186035, "num_tokens": 20533758.0, "step": 540 }, { "epoch": 0.06882076071746597, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.978250980377197, "learning_rate": 2.2891055532005085e-07, "loss": 0.6054, "mean_token_accuracy": 0.8133206963539124, "num_tokens": 20567259.0, "step": 541 }, { "epoch": 0.06894797099605648, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 5.0493083000183105, "learning_rate": 2.2933446375582875e-07, "loss": 0.5862, "mean_token_accuracy": 0.8195580244064331, "num_tokens": 20599333.0, "step": 542 }, { "epoch": 0.06907518127464699, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 4.895230293273926, "learning_rate": 2.297583721916066e-07, "loss": 0.5235, "mean_token_accuracy": 0.8352952003479004, "num_tokens": 20640195.0, "step": 543 }, { "epoch": 0.0692023915532375, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.86189603805542, "learning_rate": 2.301822806273845e-07, "loss": 0.5441, "mean_token_accuracy": 0.8288892507553101, "num_tokens": 20682617.0, "step": 544 }, { "epoch": 0.06932960183182801, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 5.157252788543701, "learning_rate": 2.3060618906316234e-07, "loss": 0.5961, "mean_token_accuracy": 0.8173173069953918, "num_tokens": 20718032.0, "step": 545 }, { "epoch": 0.06945681211041853, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.967914581298828, "learning_rate": 2.3103009749894024e-07, "loss": 0.5604, "mean_token_accuracy": 0.8242443799972534, "num_tokens": 20755190.0, "step": 546 }, { "epoch": 0.06958402238900903, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 4.565735340118408, "learning_rate": 2.3145400593471808e-07, "loss": 0.5853, "mean_token_accuracy": 0.8159381151199341, "num_tokens": 20799088.0, "step": 547 }, { "epoch": 0.06971123266759954, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 1.9371509552001953e-06, "grad_norm": 5.034348964691162, "learning_rate": 2.3187791437049598e-07, "loss": 0.6161, "mean_token_accuracy": 0.8064411878585815, "num_tokens": 20831819.0, "step": 548 }, { "epoch": 0.06983844294619006, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 5.258508682250977, "learning_rate": 2.3230182280627383e-07, "loss": 0.5455, "mean_token_accuracy": 0.828586757183075, "num_tokens": 20861291.0, "step": 549 }, { "epoch": 0.06996565322478056, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.808994770050049, "learning_rate": 2.327257312420517e-07, "loss": 0.5224, "mean_token_accuracy": 0.8366875648498535, "num_tokens": 20897446.0, "step": 550 }, { "epoch": 0.07009286350337107, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.295971870422363, "learning_rate": 2.3314963967782957e-07, "loss": 0.5454, "mean_token_accuracy": 0.8244921565055847, "num_tokens": 20934423.0, "step": 551 }, { "epoch": 0.07022007378196159, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.9166717529296875, "learning_rate": 2.3357354811360745e-07, "loss": 0.5635, "mean_token_accuracy": 0.8257346153259277, "num_tokens": 20978394.0, "step": 552 }, { "epoch": 0.07034728406055209, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.217577934265137, "learning_rate": 2.3399745654938532e-07, "loss": 0.5893, "mean_token_accuracy": 0.8169148564338684, "num_tokens": 21010669.0, "step": 553 }, { "epoch": 0.0704744943391426, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.784279823303223, "learning_rate": 2.344213649851632e-07, "loss": 0.5587, "mean_token_accuracy": 0.8271930813789368, "num_tokens": 21047696.0, "step": 554 }, { "epoch": 0.07060170461773312, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.291534423828125e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.806813716888428, "learning_rate": 2.3484527342094106e-07, "loss": 0.6398, "mean_token_accuracy": 0.8040369749069214, "num_tokens": 21091437.0, "step": 555 }, { "epoch": 0.07072891489632362, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.077147960662842, "learning_rate": 2.3526918185671894e-07, "loss": 0.6007, "mean_token_accuracy": 0.8129279017448425, "num_tokens": 21128880.0, "step": 556 }, { "epoch": 0.07085612517491413, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.903563499450684, "learning_rate": 2.356930902924968e-07, "loss": 0.5524, "mean_token_accuracy": 0.827279806137085, "num_tokens": 21168523.0, "step": 557 }, { "epoch": 0.07098333545350465, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.9014973640441895, "learning_rate": 2.3611699872827468e-07, "loss": 0.6385, "mean_token_accuracy": 0.8007320165634155, "num_tokens": 21211376.0, "step": 558 }, { "epoch": 0.07111054573209516, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 4.782688140869141, "learning_rate": 2.3654090716405255e-07, "loss": 0.5711, "mean_token_accuracy": 0.8215511441230774, "num_tokens": 21249162.0, "step": 559 }, { "epoch": 0.07123775601068566, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 1.9669532775878906e-06, "grad_norm": 5.017435073852539, "learning_rate": 2.3696481559983043e-07, "loss": 0.598, "mean_token_accuracy": 0.8120740056037903, "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": 4.975734233856201, "learning_rate": 2.373887240356083e-07, "loss": 0.5885, "mean_token_accuracy": 0.8168797492980957, "num_tokens": 21319508.0, "step": 561 }, { "epoch": 0.07149217656786669, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.955165863037109, "learning_rate": 2.3781263247138617e-07, "loss": 0.553, "mean_token_accuracy": 0.827865719795227, "num_tokens": 21357111.0, "step": 562 }, { "epoch": 0.07161938684645719, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 5.034229755401611, "learning_rate": 2.3823654090716404e-07, "loss": 0.5546, "mean_token_accuracy": 0.8225544691085815, "num_tokens": 21393222.0, "step": 563 }, { "epoch": 0.0717465971250477, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 1.9818544387817383e-06, "grad_norm": 4.860071182250977, "learning_rate": 2.386604493429419e-07, "loss": 0.5098, "mean_token_accuracy": 0.8366718292236328, "num_tokens": 21429595.0, "step": 564 }, { "epoch": 0.07187380740363822, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.973532676696777, "learning_rate": 2.390843577787198e-07, "loss": 0.6051, "mean_token_accuracy": 0.807648777961731, "num_tokens": 21467341.0, "step": 565 }, { "epoch": 0.07200101768222872, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.785694122314453, "learning_rate": 2.3950826621449766e-07, "loss": 0.483, "mean_token_accuracy": 0.8480539917945862, "num_tokens": 21510927.0, "step": 566 }, { "epoch": 0.07212822796081923, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.862033367156982, "learning_rate": 2.3993217465027556e-07, "loss": 0.5586, "mean_token_accuracy": 0.8252994418144226, "num_tokens": 21551604.0, "step": 567 }, { "epoch": 0.07225543823940975, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 5.128532409667969, "learning_rate": 2.403560830860534e-07, "loss": 0.6091, "mean_token_accuracy": 0.813651978969574, "num_tokens": 21591871.0, "step": 568 }, { "epoch": 0.07238264851800025, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 5.072410583496094, "learning_rate": 2.4077999152183125e-07, "loss": 0.6066, "mean_token_accuracy": 0.8099791407585144, "num_tokens": 21629810.0, "step": 569 }, { "epoch": 0.07250985879659076, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 5.060544490814209, "learning_rate": 2.4120389995760915e-07, "loss": 0.5052, "mean_token_accuracy": 0.8431170582771301, "num_tokens": 21668706.0, "step": 570 }, { "epoch": 0.07263706907518128, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 4.913941860198975, "learning_rate": 2.41627808393387e-07, "loss": 0.5493, "mean_token_accuracy": 0.8250609636306763, "num_tokens": 21712725.0, "step": 571 }, { "epoch": 0.07276427935377179, "ewc_loss": 0.00244140625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 4.8499016761779785, "learning_rate": 2.420517168291649e-07, "loss": 0.5592, "mean_token_accuracy": 0.8271862268447876, "num_tokens": 21746086.0, "step": 572 }, { "epoch": 0.07289148963236229, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.892177104949951, "learning_rate": 2.4247562526494274e-07, "loss": 0.5192, "mean_token_accuracy": 0.8335530757904053, "num_tokens": 21779573.0, "step": 573 }, { "epoch": 0.0730186999109528, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.8216400146484375, "learning_rate": 2.4289953370072064e-07, "loss": 0.5915, "mean_token_accuracy": 0.8143519163131714, "num_tokens": 21818766.0, "step": 574 }, { "epoch": 0.07314591018954332, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 4.832194805145264, "learning_rate": 2.433234421364985e-07, "loss": 0.547, "mean_token_accuracy": 0.8322341442108154, "num_tokens": 21855539.0, "step": 575 }, { "epoch": 0.07327312046813382, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 4.806090831756592, "learning_rate": 2.437473505722764e-07, "loss": 0.5125, "mean_token_accuracy": 0.8360037803649902, "num_tokens": 21892567.0, "step": 576 }, { "epoch": 0.07340033074672433, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 4.835212230682373, "learning_rate": 2.4417125900805423e-07, "loss": 0.5311, "mean_token_accuracy": 0.8254250288009644, "num_tokens": 21928607.0, "step": 577 }, { "epoch": 0.07352754102531485, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.92312479019165, "learning_rate": 2.4459516744383213e-07, "loss": 0.5783, "mean_token_accuracy": 0.8193802833557129, "num_tokens": 21963781.0, "step": 578 }, { "epoch": 0.07365475130390535, "ewc_loss": 0.002532958984375, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.957513332366943, "learning_rate": 2.4501907587961e-07, "loss": 0.5387, "mean_token_accuracy": 0.8313257694244385, "num_tokens": 21999140.0, "step": 579 }, { "epoch": 0.07378196158249586, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 4.649156093597412, "learning_rate": 2.454429843153879e-07, "loss": 0.6377, "mean_token_accuracy": 0.8006879091262817, "num_tokens": 22038703.0, "step": 580 }, { "epoch": 0.07390917186108638, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 5.070023536682129, "learning_rate": 2.458668927511657e-07, "loss": 0.5524, "mean_token_accuracy": 0.8191067576408386, "num_tokens": 22071052.0, "step": 581 }, { "epoch": 0.07403638213967688, "ewc_loss": 0.002532958984375, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.976696491241455, "learning_rate": 2.462908011869436e-07, "loss": 0.5176, "mean_token_accuracy": 0.8366468548774719, "num_tokens": 22109629.0, "step": 582 }, { "epoch": 0.0741635924182674, "ewc_loss": 0.002532958984375, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.743562698364258, "learning_rate": 2.4671470962272147e-07, "loss": 0.5295, "mean_token_accuracy": 0.8339696526527405, "num_tokens": 22151544.0, "step": 583 }, { "epoch": 0.07429080269685791, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 5.0135393142700195, "learning_rate": 2.4713861805849937e-07, "loss": 0.5075, "mean_token_accuracy": 0.8362327814102173, "num_tokens": 22189241.0, "step": 584 }, { "epoch": 0.07441801297544842, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 4.779636859893799, "learning_rate": 2.475625264942772e-07, "loss": 0.5229, "mean_token_accuracy": 0.8342139720916748, "num_tokens": 22229138.0, "step": 585 }, { "epoch": 0.07454522325403892, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 4.808167457580566, "learning_rate": 2.479864349300551e-07, "loss": 0.561, "mean_token_accuracy": 0.825167715549469, "num_tokens": 22264930.0, "step": 586 }, { "epoch": 0.07467243353262944, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 4.889814376831055, "learning_rate": 2.4841034336583296e-07, "loss": 0.5033, "mean_token_accuracy": 0.8350579142570496, "num_tokens": 22297162.0, "step": 587 }, { "epoch": 0.07479964381121995, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 4.808210372924805, "learning_rate": 2.488342518016108e-07, "loss": 0.5866, "mean_token_accuracy": 0.8128963112831116, "num_tokens": 22334779.0, "step": 588 }, { "epoch": 0.07492685408981045, "ewc_loss": 0.0025177001953125, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 4.858711242675781, "learning_rate": 2.492581602373887e-07, "loss": 0.544, "mean_token_accuracy": 0.8306999206542969, "num_tokens": 22372820.0, "step": 589 }, { "epoch": 0.07505406436840097, "ewc_loss": 0.002532958984375, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.853765487670898, "learning_rate": 2.4968206867316655e-07, "loss": 0.5641, "mean_token_accuracy": 0.8223062753677368, "num_tokens": 22417841.0, "step": 590 }, { "epoch": 0.07518127464699148, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 4.871220111846924, "learning_rate": 2.5010597710894445e-07, "loss": 0.479, "mean_token_accuracy": 0.8476495742797852, "num_tokens": 22449133.0, "step": 591 }, { "epoch": 0.07530848492558198, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 4.847168445587158, "learning_rate": 2.505298855447223e-07, "loss": 0.5438, "mean_token_accuracy": 0.8288641571998596, "num_tokens": 22491145.0, "step": 592 }, { "epoch": 0.0754356952041725, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.840909004211426, "learning_rate": 2.509537939805002e-07, "loss": 0.5566, "mean_token_accuracy": 0.8274878263473511, "num_tokens": 22531738.0, "step": 593 }, { "epoch": 0.07556290548276301, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 4.649162292480469e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 5.110917091369629, "learning_rate": 2.513777024162781e-07, "loss": 0.5355, "mean_token_accuracy": 0.8326998353004456, "num_tokens": 22563005.0, "step": 594 }, { "epoch": 0.07569011576135352, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.778317928314209, "learning_rate": 2.5180161085205594e-07, "loss": 0.5171, "mean_token_accuracy": 0.8376683592796326, "num_tokens": 22601333.0, "step": 595 }, { "epoch": 0.07581732603994402, "ewc_loss": 0.002532958984375, "ewc_loss_diag": 4.649162292480469e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 4.851387977600098, "learning_rate": 2.522255192878338e-07, "loss": 0.5704, "mean_token_accuracy": 0.8203060030937195, "num_tokens": 22639528.0, "step": 596 }, { "epoch": 0.07594453631853454, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 4.649162292480469e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 5.071884632110596, "learning_rate": 2.526494277236117e-07, "loss": 0.5542, "mean_token_accuracy": 0.8255683183670044, "num_tokens": 22675290.0, "step": 597 }, { "epoch": 0.07607174659712505, "ewc_loss": 0.0026092529296875, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 5.149494647979736, "learning_rate": 2.530733361593896e-07, "loss": 0.5288, "mean_token_accuracy": 0.8313812017440796, "num_tokens": 22712823.0, "step": 598 }, { "epoch": 0.07619895687571555, "ewc_loss": 0.0026092529296875, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 4.965847492218018, "learning_rate": 2.5349724459516743e-07, "loss": 0.5507, "mean_token_accuracy": 0.8270182013511658, "num_tokens": 22750916.0, "step": 599 }, { "epoch": 0.07632616715430607, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.907285213470459, "learning_rate": 2.539211530309453e-07, "loss": 0.5021, "mean_token_accuracy": 0.8381158113479614, "num_tokens": 22782537.0, "step": 600 }, { "epoch": 0.07645337743289658, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 5.108915328979492, "learning_rate": 2.543450614667232e-07, "loss": 0.5934, "mean_token_accuracy": 0.8155394792556763, "num_tokens": 22814805.0, "step": 601 }, { "epoch": 0.07658058771148708, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 5.011478424072266, "learning_rate": 2.547689699025011e-07, "loss": 0.5918, "mean_token_accuracy": 0.8134022355079651, "num_tokens": 22852654.0, "step": 602 }, { "epoch": 0.0767077979900776, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.890590190887451, "learning_rate": 2.551928783382789e-07, "loss": 0.5459, "mean_token_accuracy": 0.82569420337677, "num_tokens": 22894205.0, "step": 603 }, { "epoch": 0.07683500826866811, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 4.91323709487915, "learning_rate": 2.5561678677405677e-07, "loss": 0.5124, "mean_token_accuracy": 0.8348146677017212, "num_tokens": 22924095.0, "step": 604 }, { "epoch": 0.07696221854725861, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 4.869500160217285, "learning_rate": 2.5604069520983467e-07, "loss": 0.5946, "mean_token_accuracy": 0.8108688592910767, "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": 4.802460670471191, "learning_rate": 2.564646036456125e-07, "loss": 0.5466, "mean_token_accuracy": 0.8249140381813049, "num_tokens": 23004670.0, "step": 606 }, { "epoch": 0.07721663910443964, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 5.007536888122559, "learning_rate": 2.568885120813904e-07, "loss": 0.5468, "mean_token_accuracy": 0.8244079947471619, "num_tokens": 23041999.0, "step": 607 }, { "epoch": 0.07734384938303016, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 4.915132522583008, "learning_rate": 2.5731242051716826e-07, "loss": 0.5355, "mean_token_accuracy": 0.8325252532958984, "num_tokens": 23080704.0, "step": 608 }, { "epoch": 0.07747105966162066, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.965963840484619, "learning_rate": 2.5773632895294616e-07, "loss": 0.5707, "mean_token_accuracy": 0.8214924335479736, "num_tokens": 23118662.0, "step": 609 }, { "epoch": 0.07759826994021117, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 4.873659133911133, "learning_rate": 2.58160237388724e-07, "loss": 0.4935, "mean_token_accuracy": 0.8395394086837769, "num_tokens": 23150367.0, "step": 610 }, { "epoch": 0.07772548021880168, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.930520057678223, "learning_rate": 2.585841458245019e-07, "loss": 0.5517, "mean_token_accuracy": 0.8254759311676025, "num_tokens": 23185836.0, "step": 611 }, { "epoch": 0.07785269049739219, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.921849250793457, "learning_rate": 2.5900805426027975e-07, "loss": 0.4926, "mean_token_accuracy": 0.8419750332832336, "num_tokens": 23219346.0, "step": 612 }, { "epoch": 0.0779799007759827, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.947061061859131, "learning_rate": 2.5943196269605765e-07, "loss": 0.5637, "mean_token_accuracy": 0.8277204036712646, "num_tokens": 23257569.0, "step": 613 }, { "epoch": 0.07810711105457321, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.710419178009033, "learning_rate": 2.598558711318355e-07, "loss": 0.5543, "mean_token_accuracy": 0.8275665044784546, "num_tokens": 23302944.0, "step": 614 }, { "epoch": 0.07823432133316371, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 4.928945064544678, "learning_rate": 2.602797795676134e-07, "loss": 0.5603, "mean_token_accuracy": 0.8234522342681885, "num_tokens": 23341660.0, "step": 615 }, { "epoch": 0.07836153161175423, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 4.995841026306152, "learning_rate": 2.6070368800339124e-07, "loss": 0.5576, "mean_token_accuracy": 0.8188800811767578, "num_tokens": 23375956.0, "step": 616 }, { "epoch": 0.07848874189034474, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 4.852602481842041, "learning_rate": 2.6112759643916914e-07, "loss": 0.5098, "mean_token_accuracy": 0.8361591100692749, "num_tokens": 23407568.0, "step": 617 }, { "epoch": 0.07861595216893524, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 4.9517998695373535, "learning_rate": 2.61551504874947e-07, "loss": 0.5425, "mean_token_accuracy": 0.8299029469490051, "num_tokens": 23440261.0, "step": 618 }, { "epoch": 0.07874316244752576, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 4.958657741546631, "learning_rate": 2.619754133107249e-07, "loss": 0.5252, "mean_token_accuracy": 0.8347262740135193, "num_tokens": 23481706.0, "step": 619 }, { "epoch": 0.07887037272611627, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 5.027273654937744, "learning_rate": 2.623993217465028e-07, "loss": 0.5853, "mean_token_accuracy": 0.8128696084022522, "num_tokens": 23515410.0, "step": 620 }, { "epoch": 0.07899758300470679, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.798173904418945e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.816722869873047, "learning_rate": 2.6282323018228063e-07, "loss": 0.5691, "mean_token_accuracy": 0.8241187334060669, "num_tokens": 23554627.0, "step": 621 }, { "epoch": 0.07912479328329729, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 4.827976226806641e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 4.856637001037598, "learning_rate": 2.632471386180585e-07, "loss": 0.5922, "mean_token_accuracy": 0.8129410743713379, "num_tokens": 23593400.0, "step": 622 }, { "epoch": 0.0792520035618878, "ewc_loss": 0.0026702880859375, "ewc_loss_diag": 4.827976226806641e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 5.094511985778809, "learning_rate": 2.6367104705383637e-07, "loss": 0.5899, "mean_token_accuracy": 0.813047468662262, "num_tokens": 23628242.0, "step": 623 }, { "epoch": 0.07937921384047832, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.827976226806641e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 4.948005199432373, "learning_rate": 2.6409495548961427e-07, "loss": 0.5761, "mean_token_accuracy": 0.8235859870910645, "num_tokens": 23671041.0, "step": 624 }, { "epoch": 0.07950642411906882, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 5.060888290405273, "learning_rate": 2.6451886392539206e-07, "loss": 0.5533, "mean_token_accuracy": 0.8283077478408813, "num_tokens": 23701904.0, "step": 625 }, { "epoch": 0.07963363439765933, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.829959392547607, "learning_rate": 2.6494277236116996e-07, "loss": 0.5174, "mean_token_accuracy": 0.8351616263389587, "num_tokens": 23742374.0, "step": 626 }, { "epoch": 0.07976084467624985, "ewc_loss": 0.002685546875, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 4.949707508087158, "learning_rate": 2.6536668079694786e-07, "loss": 0.6227, "mean_token_accuracy": 0.806579053401947, "num_tokens": 23780216.0, "step": 627 }, { "epoch": 0.07988805495484035, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 5.225978374481201, "learning_rate": 2.6579058923272576e-07, "loss": 0.4772, "mean_token_accuracy": 0.8495505452156067, "num_tokens": 23813641.0, "step": 628 }, { "epoch": 0.08001526523343086, "ewc_loss": 0.0027618408203125, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 4.910791873931885, "learning_rate": 2.6621449766850356e-07, "loss": 0.5621, "mean_token_accuracy": 0.8246006965637207, "num_tokens": 23850189.0, "step": 629 }, { "epoch": 0.08014247551202137, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.851794242858887, "learning_rate": 2.6663840610428145e-07, "loss": 0.5163, "mean_token_accuracy": 0.8354261517524719, "num_tokens": 23884410.0, "step": 630 }, { "epoch": 0.08026968579061187, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.0067901611328125e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 4.839031219482422, "learning_rate": 2.6706231454005935e-07, "loss": 0.5262, "mean_token_accuracy": 0.8338105082511902, "num_tokens": 23930139.0, "step": 631 }, { "epoch": 0.08039689606920239, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.715879917144775, "learning_rate": 2.6748622297583725e-07, "loss": 0.5245, "mean_token_accuracy": 0.8342233896255493, "num_tokens": 23973124.0, "step": 632 }, { "epoch": 0.0805241063477929, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 5.0067901611328125e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 4.85662841796875, "learning_rate": 2.6791013141161505e-07, "loss": 0.5328, "mean_token_accuracy": 0.8331080079078674, "num_tokens": 24008676.0, "step": 633 }, { "epoch": 0.08065131662638342, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 4.976987838745117e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 5.032565593719482, "learning_rate": 2.6833403984739294e-07, "loss": 0.5316, "mean_token_accuracy": 0.8315946459770203, "num_tokens": 24044880.0, "step": 634 }, { "epoch": 0.08077852690497392, "ewc_loss": 0.0027618408203125, "ewc_loss_diag": 5.0067901611328125e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.82649040222168, "learning_rate": 2.6875794828317084e-07, "loss": 0.522, "mean_token_accuracy": 0.8380435705184937, "num_tokens": 24090140.0, "step": 635 }, { "epoch": 0.08090573718356443, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.0067901611328125e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 4.992520809173584, "learning_rate": 2.6918185671894874e-07, "loss": 0.6065, "mean_token_accuracy": 0.8139027953147888, "num_tokens": 24123311.0, "step": 636 }, { "epoch": 0.08103294746215495, "ewc_loss": 0.0027618408203125, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.949027061462402, "learning_rate": 2.6960576515472654e-07, "loss": 0.6403, "mean_token_accuracy": 0.7990856170654297, "num_tokens": 24164019.0, "step": 637 }, { "epoch": 0.08116015774074545, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 2.250075340270996e-06, "grad_norm": 4.936344623565674, "learning_rate": 2.7002967359050443e-07, "loss": 0.5075, "mean_token_accuracy": 0.8393823504447937, "num_tokens": 24205986.0, "step": 638 }, { "epoch": 0.08128736801933596, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 4.825570583343506, "learning_rate": 2.7045358202628233e-07, "loss": 0.5643, "mean_token_accuracy": 0.8186653852462769, "num_tokens": 24251956.0, "step": 639 }, { "epoch": 0.08141457829792648, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 5.066394805908203e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 4.887601852416992, "learning_rate": 2.7087749046206023e-07, "loss": 0.5106, "mean_token_accuracy": 0.8359513282775879, "num_tokens": 24284990.0, "step": 640 }, { "epoch": 0.08154178857651698, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.066394805908203e-06, "ewc_loss_parallel": 2.250075340270996e-06, "grad_norm": 5.007438659667969, "learning_rate": 2.71301398897838e-07, "loss": 0.5665, "mean_token_accuracy": 0.8208193778991699, "num_tokens": 24323305.0, "step": 641 }, { "epoch": 0.08166899885510749, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.8901567459106445, "learning_rate": 2.717253073336159e-07, "loss": 0.4785, "mean_token_accuracy": 0.8463770151138306, "num_tokens": 24365847.0, "step": 642 }, { "epoch": 0.081796209133698, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.930727005004883, "learning_rate": 2.721492157693938e-07, "loss": 0.4851, "mean_token_accuracy": 0.8451962471008301, "num_tokens": 24407269.0, "step": 643 }, { "epoch": 0.0819234194122885, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 5.125999450683594e-06, "ewc_loss_parallel": 2.250075340270996e-06, "grad_norm": 4.919972896575928, "learning_rate": 2.7257312420517167e-07, "loss": 0.5416, "mean_token_accuracy": 0.8277755975723267, "num_tokens": 24444990.0, "step": 644 }, { "epoch": 0.08205062969087902, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.125999450683594e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 4.81396484375, "learning_rate": 2.729970326409495e-07, "loss": 0.5719, "mean_token_accuracy": 0.8225423693656921, "num_tokens": 24487293.0, "step": 645 }, { "epoch": 0.08217783996946953, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 5.011653900146484, "learning_rate": 2.734209410767274e-07, "loss": 0.534, "mean_token_accuracy": 0.8307636976242065, "num_tokens": 24521403.0, "step": 646 }, { "epoch": 0.08230505024806005, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 4.984034538269043, "learning_rate": 2.738448495125053e-07, "loss": 0.5434, "mean_token_accuracy": 0.8308174014091492, "num_tokens": 24562740.0, "step": 647 }, { "epoch": 0.08243226052665055, "ewc_loss": 0.0027923583984375, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 4.906425952911377, "learning_rate": 2.7426875794828316e-07, "loss": 0.5504, "mean_token_accuracy": 0.821962833404541, "num_tokens": 24597507.0, "step": 648 }, { "epoch": 0.08255947080524106, "ewc_loss": 0.0027923583984375, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 4.860665798187256, "learning_rate": 2.74692666384061e-07, "loss": 0.5118, "mean_token_accuracy": 0.8354684710502625, "num_tokens": 24635629.0, "step": 649 }, { "epoch": 0.08268668108383158, "ewc_loss": 0.0027923583984375, "ewc_loss_diag": 5.21540641784668e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 4.887802600860596, "learning_rate": 2.751165748198389e-07, "loss": 0.5381, "mean_token_accuracy": 0.8282355070114136, "num_tokens": 24676087.0, "step": 650 }, { "epoch": 0.08281389136242208, "ewc_loss": 0.0028228759765625, "ewc_loss_diag": 5.21540641784668e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 5.0723371505737305, "learning_rate": 2.755404832556168e-07, "loss": 0.556, "mean_token_accuracy": 0.8244994878768921, "num_tokens": 24713652.0, "step": 651 }, { "epoch": 0.0829411016410126, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 4.916191577911377, "learning_rate": 2.7596439169139465e-07, "loss": 0.5385, "mean_token_accuracy": 0.8303972482681274, "num_tokens": 24750130.0, "step": 652 }, { "epoch": 0.08306831191960311, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 4.968290328979492, "learning_rate": 2.763883001271725e-07, "loss": 0.588, "mean_token_accuracy": 0.8146637678146362, "num_tokens": 24787337.0, "step": 653 }, { "epoch": 0.08319552219819361, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 5.014115333557129, "learning_rate": 2.768122085629504e-07, "loss": 0.5799, "mean_token_accuracy": 0.811046838760376, "num_tokens": 24822736.0, "step": 654 }, { "epoch": 0.08332273247678412, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 2.339482307434082e-06, "grad_norm": 5.017099857330322, "learning_rate": 2.772361169987283e-07, "loss": 0.5215, "mean_token_accuracy": 0.8377408385276794, "num_tokens": 24858743.0, "step": 655 }, { "epoch": 0.08344994275537464, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 4.90596866607666, "learning_rate": 2.7766002543450614e-07, "loss": 0.5296, "mean_token_accuracy": 0.8291728496551514, "num_tokens": 24901286.0, "step": 656 }, { "epoch": 0.08357715303396514, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 4.882938861846924, "learning_rate": 2.78083933870284e-07, "loss": 0.5991, "mean_token_accuracy": 0.8139235377311707, "num_tokens": 24943457.0, "step": 657 }, { "epoch": 0.08370436331255565, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 4.976430892944336, "learning_rate": 2.785078423060619e-07, "loss": 0.5388, "mean_token_accuracy": 0.826244056224823, "num_tokens": 24979247.0, "step": 658 }, { "epoch": 0.08383157359114617, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 2.3543834686279297e-06, "grad_norm": 5.03066873550415, "learning_rate": 2.789317507418398e-07, "loss": 0.5512, "mean_token_accuracy": 0.8241589665412903, "num_tokens": 25017456.0, "step": 659 }, { "epoch": 0.08395878386973668, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.304813385009766e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.929226875305176, "learning_rate": 2.7935565917761763e-07, "loss": 0.5079, "mean_token_accuracy": 0.8363699913024902, "num_tokens": 25054682.0, "step": 660 }, { "epoch": 0.08408599414832718, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 2.339482307434082e-06, "grad_norm": 5.107802391052246, "learning_rate": 2.797795676133955e-07, "loss": 0.5811, "mean_token_accuracy": 0.8153480887413025, "num_tokens": 25092501.0, "step": 661 }, { "epoch": 0.0842132044269177, "ewc_loss": 0.0029144287109375, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 4.9733991622924805, "learning_rate": 2.802034760491734e-07, "loss": 0.4496, "mean_token_accuracy": 0.8559743165969849, "num_tokens": 25129253.0, "step": 662 }, { "epoch": 0.08434041470550821, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.802022457122803, "learning_rate": 2.806273844849512e-07, "loss": 0.4888, "mean_token_accuracy": 0.8441116809844971, "num_tokens": 25169119.0, "step": 663 }, { "epoch": 0.08446762498409871, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 4.949052810668945, "learning_rate": 2.810512929207291e-07, "loss": 0.5588, "mean_token_accuracy": 0.8183895945549011, "num_tokens": 25207654.0, "step": 664 }, { "epoch": 0.08459483526268922, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 4.930335521697998, "learning_rate": 2.8147520135650697e-07, "loss": 0.5042, "mean_token_accuracy": 0.8365657329559326, "num_tokens": 25243258.0, "step": 665 }, { "epoch": 0.08472204554127974, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 4.873802185058594, "learning_rate": 2.8189910979228487e-07, "loss": 0.5438, "mean_token_accuracy": 0.829972505569458, "num_tokens": 25281301.0, "step": 666 }, { "epoch": 0.08484925581987024, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.3543834686279297e-06, "grad_norm": 4.999887943267822, "learning_rate": 2.823230182280627e-07, "loss": 0.4954, "mean_token_accuracy": 0.8427292108535767, "num_tokens": 25313500.0, "step": 667 }, { "epoch": 0.08497646609846075, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 4.937926292419434, "learning_rate": 2.827469266638406e-07, "loss": 0.5323, "mean_token_accuracy": 0.8297631144523621, "num_tokens": 25355121.0, "step": 668 }, { "epoch": 0.08510367637705127, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 4.884327411651611, "learning_rate": 2.8317083509961846e-07, "loss": 0.5482, "mean_token_accuracy": 0.8282455205917358, "num_tokens": 25398868.0, "step": 669 }, { "epoch": 0.08523088665564178, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 4.87260627746582, "learning_rate": 2.8359474353539636e-07, "loss": 0.5271, "mean_token_accuracy": 0.8314221501350403, "num_tokens": 25438670.0, "step": 670 }, { "epoch": 0.08535809693423228, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 5.001886367797852, "learning_rate": 2.840186519711742e-07, "loss": 0.5735, "mean_token_accuracy": 0.8222348093986511, "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": 4.975941181182861, "learning_rate": 2.844425604069521e-07, "loss": 0.5578, "mean_token_accuracy": 0.8214861154556274, "num_tokens": 25514615.0, "step": 672 }, { "epoch": 0.08561251749141331, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 5.027429103851318, "learning_rate": 2.8486646884272995e-07, "loss": 0.5346, "mean_token_accuracy": 0.8291518688201904, "num_tokens": 25552753.0, "step": 673 }, { "epoch": 0.08573972777000381, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 4.997378349304199, "learning_rate": 2.8529037727850785e-07, "loss": 0.4721, "mean_token_accuracy": 0.8466674089431763, "num_tokens": 25587415.0, "step": 674 }, { "epoch": 0.08586693804859433, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 4.932326793670654, "learning_rate": 2.857142857142857e-07, "loss": 0.5019, "mean_token_accuracy": 0.841447114944458, "num_tokens": 25625962.0, "step": 675 }, { "epoch": 0.08599414832718484, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 5.093207359313965, "learning_rate": 2.861381941500636e-07, "loss": 0.5503, "mean_token_accuracy": 0.8252423405647278, "num_tokens": 25669680.0, "step": 676 }, { "epoch": 0.08612135860577534, "ewc_loss": 0.0029754638671875, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 5.078682899475098, "learning_rate": 2.8656210258584144e-07, "loss": 0.5847, "mean_token_accuracy": 0.8124721646308899, "num_tokens": 25709221.0, "step": 677 }, { "epoch": 0.08624856888436586, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 5.195337772369385, "learning_rate": 2.869860110216193e-07, "loss": 0.586, "mean_token_accuracy": 0.8122925162315369, "num_tokens": 25741880.0, "step": 678 }, { "epoch": 0.08637577916295637, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.909136772155762, "learning_rate": 2.874099194573972e-07, "loss": 0.5187, "mean_token_accuracy": 0.8346798419952393, "num_tokens": 25786282.0, "step": 679 }, { "epoch": 0.08650298944154687, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 5.010886192321777, "learning_rate": 2.878338278931751e-07, "loss": 0.5646, "mean_token_accuracy": 0.822827160358429, "num_tokens": 25820229.0, "step": 680 }, { "epoch": 0.08663019972013739, "ewc_loss": 0.0029754638671875, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 4.939535140991211, "learning_rate": 2.8825773632895293e-07, "loss": 0.4999, "mean_token_accuracy": 0.8443822860717773, "num_tokens": 25856465.0, "step": 681 }, { "epoch": 0.0867574099987279, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 4.928927421569824, "learning_rate": 2.886816447647308e-07, "loss": 0.4985, "mean_token_accuracy": 0.8403347730636597, "num_tokens": 25897407.0, "step": 682 }, { "epoch": 0.08688462027731841, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.971181392669678, "learning_rate": 2.891055532005087e-07, "loss": 0.522, "mean_token_accuracy": 0.834641695022583, "num_tokens": 25938825.0, "step": 683 }, { "epoch": 0.08701183055590891, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.662441253662109e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 5.092092990875244, "learning_rate": 2.8952946163628657e-07, "loss": 0.5522, "mean_token_accuracy": 0.8219520449638367, "num_tokens": 25974944.0, "step": 684 }, { "epoch": 0.08713904083449943, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.021801948547363, "learning_rate": 2.899533700720644e-07, "loss": 0.511, "mean_token_accuracy": 0.8349251747131348, "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": 4.955615520477295, "learning_rate": 2.9037727850784227e-07, "loss": 0.5565, "mean_token_accuracy": 0.8234270811080933, "num_tokens": 26049779.0, "step": 686 }, { "epoch": 0.08739346139168044, "ewc_loss": 0.0030059814453125, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.946086406707764, "learning_rate": 2.9080118694362016e-07, "loss": 0.5274, "mean_token_accuracy": 0.832491397857666, "num_tokens": 26089095.0, "step": 687 }, { "epoch": 0.08752067167027096, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 4.9026408195495605, "learning_rate": 2.9122509537939806e-07, "loss": 0.5267, "mean_token_accuracy": 0.8328570127487183, "num_tokens": 26132553.0, "step": 688 }, { "epoch": 0.08764788194886147, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.7220458984375e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 4.974887371063232, "learning_rate": 2.916490038151759e-07, "loss": 0.5779, "mean_token_accuracy": 0.8142252564430237, "num_tokens": 26171301.0, "step": 689 }, { "epoch": 0.08777509222745197, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.751848220825195e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.021276950836182, "learning_rate": 2.9207291225095376e-07, "loss": 0.5646, "mean_token_accuracy": 0.8196811676025391, "num_tokens": 26217871.0, "step": 690 }, { "epoch": 0.08790230250604249, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 4.915095806121826, "learning_rate": 2.9249682068673166e-07, "loss": 0.5285, "mean_token_accuracy": 0.8302507400512695, "num_tokens": 26255927.0, "step": 691 }, { "epoch": 0.088029512784633, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.002311706542969, "learning_rate": 2.9292072912250955e-07, "loss": 0.5115, "mean_token_accuracy": 0.8386390209197998, "num_tokens": 26296041.0, "step": 692 }, { "epoch": 0.0881567230632235, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.085772514343262, "learning_rate": 2.933446375582874e-07, "loss": 0.5935, "mean_token_accuracy": 0.813586950302124, "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": 5.006712436676025, "learning_rate": 2.9376854599406525e-07, "loss": 0.5001, "mean_token_accuracy": 0.8376882076263428, "num_tokens": 26368837.0, "step": 694 }, { "epoch": 0.08841114362040453, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 5.059540271759033, "learning_rate": 2.9419245442984315e-07, "loss": 0.575, "mean_token_accuracy": 0.8176966905593872, "num_tokens": 26402151.0, "step": 695 }, { "epoch": 0.08853835389899505, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 4.9885382652282715, "learning_rate": 2.9461636286562104e-07, "loss": 0.5279, "mean_token_accuracy": 0.8304042816162109, "num_tokens": 26435550.0, "step": 696 }, { "epoch": 0.08866556417758555, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 4.909757137298584, "learning_rate": 2.9504027130139884e-07, "loss": 0.5179, "mean_token_accuracy": 0.8357006311416626, "num_tokens": 26479299.0, "step": 697 }, { "epoch": 0.08879277445617606, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 4.968105792999268, "learning_rate": 2.9546417973717674e-07, "loss": 0.5543, "mean_token_accuracy": 0.8239553570747375, "num_tokens": 26519380.0, "step": 698 }, { "epoch": 0.08891998473476657, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 4.996092319488525, "learning_rate": 2.9588808817295464e-07, "loss": 0.4804, "mean_token_accuracy": 0.8436684608459473, "num_tokens": 26557306.0, "step": 699 }, { "epoch": 0.08904719501335707, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.108090400695801, "learning_rate": 2.9631199660873253e-07, "loss": 0.5501, "mean_token_accuracy": 0.825908899307251, "num_tokens": 26593549.0, "step": 700 }, { "epoch": 0.08917440529194759, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 4.973219871520996, "learning_rate": 2.9673590504451033e-07, "loss": 0.4753, "mean_token_accuracy": 0.8488975763320923, "num_tokens": 26629926.0, "step": 701 }, { "epoch": 0.0893016155705381, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 4.9447197914123535, "learning_rate": 2.9715981348028823e-07, "loss": 0.5074, "mean_token_accuracy": 0.8392199277877808, "num_tokens": 26668652.0, "step": 702 }, { "epoch": 0.0894288258491286, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.144484996795654, "learning_rate": 2.975837219160661e-07, "loss": 0.5693, "mean_token_accuracy": 0.8188195824623108, "num_tokens": 26708084.0, "step": 703 }, { "epoch": 0.08955603612771912, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.017318248748779, "learning_rate": 2.98007630351844e-07, "loss": 0.5355, "mean_token_accuracy": 0.8301292061805725, "num_tokens": 26750134.0, "step": 704 }, { "epoch": 0.08968324640630963, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 5.080338478088379, "learning_rate": 2.984315387876218e-07, "loss": 0.5438, "mean_token_accuracy": 0.826029896736145, "num_tokens": 26785410.0, "step": 705 }, { "epoch": 0.08981045668490013, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.105177402496338, "learning_rate": 2.988554472233997e-07, "loss": 0.548, "mean_token_accuracy": 0.8286432027816772, "num_tokens": 26825226.0, "step": 706 }, { "epoch": 0.08993766696349065, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 4.9642229080200195, "learning_rate": 2.992793556591776e-07, "loss": 0.541, "mean_token_accuracy": 0.8271198272705078, "num_tokens": 26865401.0, "step": 707 }, { "epoch": 0.09006487724208116, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 4.9811320304870605, "learning_rate": 2.997032640949555e-07, "loss": 0.5511, "mean_token_accuracy": 0.8261637687683105, "num_tokens": 26905365.0, "step": 708 }, { "epoch": 0.09019208752067168, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.9604644775390625e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.000904083251953, "learning_rate": 3.001271725307333e-07, "loss": 0.4919, "mean_token_accuracy": 0.8426241874694824, "num_tokens": 26941348.0, "step": 709 }, { "epoch": 0.09031929779926218, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.097574710845947, "learning_rate": 3.005510809665112e-07, "loss": 0.5273, "mean_token_accuracy": 0.8323304653167725, "num_tokens": 26975803.0, "step": 710 }, { "epoch": 0.09044650807785269, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.053246974945068, "learning_rate": 3.009749894022891e-07, "loss": 0.4732, "mean_token_accuracy": 0.848328709602356, "num_tokens": 27014931.0, "step": 711 }, { "epoch": 0.0905737183564432, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 4.986451148986816, "learning_rate": 3.01398897838067e-07, "loss": 0.5297, "mean_token_accuracy": 0.8333612680435181, "num_tokens": 27055505.0, "step": 712 }, { "epoch": 0.0907009286350337, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.0283331871032715, "learning_rate": 3.018228062738448e-07, "loss": 0.5572, "mean_token_accuracy": 0.8264820575714111, "num_tokens": 27092121.0, "step": 713 }, { "epoch": 0.09082813891362422, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.0050811767578125, "learning_rate": 3.022467147096227e-07, "loss": 0.4945, "mean_token_accuracy": 0.8418824672698975, "num_tokens": 27129827.0, "step": 714 }, { "epoch": 0.09095534919221474, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 4.956298828125, "learning_rate": 3.026706231454006e-07, "loss": 0.5113, "mean_token_accuracy": 0.8356912732124329, "num_tokens": 27174977.0, "step": 715 }, { "epoch": 0.09108255947080524, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 5.054305076599121, "learning_rate": 3.0309453158117844e-07, "loss": 0.5174, "mean_token_accuracy": 0.8348872661590576, "num_tokens": 27211472.0, "step": 716 }, { "epoch": 0.09120976974939575, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.118088722229004, "learning_rate": 3.035184400169563e-07, "loss": 0.5984, "mean_token_accuracy": 0.8138344883918762, "num_tokens": 27249160.0, "step": 717 }, { "epoch": 0.09133698002798626, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.164342880249023, "learning_rate": 3.039423484527342e-07, "loss": 0.5086, "mean_token_accuracy": 0.8359235525131226, "num_tokens": 27287946.0, "step": 718 }, { "epoch": 0.09146419030657676, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.0490617752075195, "learning_rate": 3.043662568885121e-07, "loss": 0.5415, "mean_token_accuracy": 0.8232325315475464, "num_tokens": 27323388.0, "step": 719 }, { "epoch": 0.09159140058516728, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.2478346824646, "learning_rate": 3.0479016532428993e-07, "loss": 0.59, "mean_token_accuracy": 0.8121000528335571, "num_tokens": 27357524.0, "step": 720 }, { "epoch": 0.0917186108637578, "ewc_loss": 0.0031585693359375, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.064533233642578, "learning_rate": 3.052140737600678e-07, "loss": 0.4966, "mean_token_accuracy": 0.8403617143630981, "num_tokens": 27392037.0, "step": 721 }, { "epoch": 0.09184582114234831, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 5.024653911590576, "learning_rate": 3.056379821958457e-07, "loss": 0.5098, "mean_token_accuracy": 0.8383665084838867, "num_tokens": 27429149.0, "step": 722 }, { "epoch": 0.09197303142093881, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.111610412597656, "learning_rate": 3.060618906316236e-07, "loss": 0.5132, "mean_token_accuracy": 0.8380769491195679, "num_tokens": 27465023.0, "step": 723 }, { "epoch": 0.09210024169952932, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5331974029541016e-06, "grad_norm": 5.097256660461426, "learning_rate": 3.064857990674014e-07, "loss": 0.5051, "mean_token_accuracy": 0.8402149677276611, "num_tokens": 27501268.0, "step": 724 }, { "epoch": 0.09222745197811984, "ewc_loss": 0.0031585693359375, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.0922322273254395, "learning_rate": 3.0690970750317927e-07, "loss": 0.5573, "mean_token_accuracy": 0.8237142562866211, "num_tokens": 27539937.0, "step": 725 }, { "epoch": 0.09235466225671034, "ewc_loss": 0.0031585693359375, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.111866474151611, "learning_rate": 3.0733361593895717e-07, "loss": 0.5194, "mean_token_accuracy": 0.8335137367248535, "num_tokens": 27574576.0, "step": 726 }, { "epoch": 0.09248187253530085, "ewc_loss": 0.0031585693359375, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 2.5480985641479492e-06, "grad_norm": 5.14680814743042, "learning_rate": 3.0775752437473507e-07, "loss": 0.5283, "mean_token_accuracy": 0.8283626437187195, "num_tokens": 27616036.0, "step": 727 }, { "epoch": 0.09260908281389137, "ewc_loss": 0.003173828125, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 4.995369911193848, "learning_rate": 3.081814328105129e-07, "loss": 0.5507, "mean_token_accuracy": 0.8248447179794312, "num_tokens": 27662040.0, "step": 728 }, { "epoch": 0.09273629309248187, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.059462070465088, "learning_rate": 3.0860534124629076e-07, "loss": 0.58, "mean_token_accuracy": 0.8163273334503174, "num_tokens": 27705136.0, "step": 729 }, { "epoch": 0.09286350337107238, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.189834117889404, "learning_rate": 3.0902924968206866e-07, "loss": 0.5297, "mean_token_accuracy": 0.8318638801574707, "num_tokens": 27740564.0, "step": 730 }, { "epoch": 0.0929907136496629, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.2848358154296875, "learning_rate": 3.0945315811784656e-07, "loss": 0.5032, "mean_token_accuracy": 0.8371846675872803, "num_tokens": 27774788.0, "step": 731 }, { "epoch": 0.0931179239282534, "ewc_loss": 0.0032196044921875, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.05783748626709, "learning_rate": 3.098770665536244e-07, "loss": 0.6475, "mean_token_accuracy": 0.7928657531738281, "num_tokens": 27818300.0, "step": 732 }, { "epoch": 0.09324513420684391, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 5.089239597320557, "learning_rate": 3.1030097498940225e-07, "loss": 0.491, "mean_token_accuracy": 0.8418307304382324, "num_tokens": 27857739.0, "step": 733 }, { "epoch": 0.09337234448543442, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.156825542449951, "learning_rate": 3.1072488342518015e-07, "loss": 0.5293, "mean_token_accuracy": 0.8256067633628845, "num_tokens": 27893897.0, "step": 734 }, { "epoch": 0.09349955476402494, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.083381652832031, "learning_rate": 3.11148791860958e-07, "loss": 0.5163, "mean_token_accuracy": 0.8327217102050781, "num_tokens": 27930511.0, "step": 735 }, { "epoch": 0.09362676504261544, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.232847690582275, "learning_rate": 3.115727002967359e-07, "loss": 0.5343, "mean_token_accuracy": 0.828771710395813, "num_tokens": 27961128.0, "step": 736 }, { "epoch": 0.09375397532120595, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.072460174560547, "learning_rate": 3.1199660873251374e-07, "loss": 0.5348, "mean_token_accuracy": 0.8312901258468628, "num_tokens": 28000748.0, "step": 737 }, { "epoch": 0.09388118559979647, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 6.22868537902832e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 5.020132064819336, "learning_rate": 3.1242051716829164e-07, "loss": 0.5681, "mean_token_accuracy": 0.8170619010925293, "num_tokens": 28046455.0, "step": 738 }, { "epoch": 0.09400839587838697, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 6.22868537902832e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 5.122952938079834, "learning_rate": 3.128444256040695e-07, "loss": 0.5442, "mean_token_accuracy": 0.8280503749847412, "num_tokens": 28087026.0, "step": 739 }, { "epoch": 0.09413560615697748, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.22868537902832e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.01823616027832, "learning_rate": 3.132683340398474e-07, "loss": 0.4672, "mean_token_accuracy": 0.8475996255874634, "num_tokens": 28126382.0, "step": 740 }, { "epoch": 0.094262816435568, "ewc_loss": 0.0032196044921875, "ewc_loss_diag": 6.22868537902832e-06, "ewc_loss_parallel": 2.592802047729492e-06, "grad_norm": 5.043061256408691, "learning_rate": 3.1369224247562523e-07, "loss": 0.4935, "mean_token_accuracy": 0.842064380645752, "num_tokens": 28161198.0, "step": 741 }, { "epoch": 0.0943900267141585, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.0557026863098145, "learning_rate": 3.1411615091140313e-07, "loss": 0.463, "mean_token_accuracy": 0.8510575294494629, "num_tokens": 28204612.0, "step": 742 }, { "epoch": 0.09451723699274901, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.258487701416016e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.062652587890625, "learning_rate": 3.14540059347181e-07, "loss": 0.5171, "mean_token_accuracy": 0.834792971611023, "num_tokens": 28241050.0, "step": 743 }, { "epoch": 0.09464444727133953, "ewc_loss": 0.0032501220703125, "ewc_loss_diag": 6.258487701416016e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.09762716293335, "learning_rate": 3.149639677829589e-07, "loss": 0.4776, "mean_token_accuracy": 0.8447462320327759, "num_tokens": 28278032.0, "step": 744 }, { "epoch": 0.09477165754993004, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 6.318092346191406e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.101864337921143, "learning_rate": 3.153878762187368e-07, "loss": 0.5642, "mean_token_accuracy": 0.8250529766082764, "num_tokens": 28315599.0, "step": 745 }, { "epoch": 0.09489886782852054, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.318092346191406e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.0987868309021, "learning_rate": 3.158117846545146e-07, "loss": 0.5883, "mean_token_accuracy": 0.8159143924713135, "num_tokens": 28358504.0, "step": 746 }, { "epoch": 0.09502607810711106, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.318092346191406e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.136375904083252, "learning_rate": 3.1623569309029247e-07, "loss": 0.5047, "mean_token_accuracy": 0.8383095264434814, "num_tokens": 28389626.0, "step": 747 }, { "epoch": 0.09515328838570157, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.033326625823975, "learning_rate": 3.1665960152607037e-07, "loss": 0.4866, "mean_token_accuracy": 0.842214822769165, "num_tokens": 28430361.0, "step": 748 }, { "epoch": 0.09528049866429207, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.060900688171387, "learning_rate": 3.1708350996184826e-07, "loss": 0.524, "mean_token_accuracy": 0.8316853642463684, "num_tokens": 28469792.0, "step": 749 }, { "epoch": 0.09540770894288259, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.136244773864746, "learning_rate": 3.175074183976261e-07, "loss": 0.5026, "mean_token_accuracy": 0.8416999578475952, "num_tokens": 28507569.0, "step": 750 }, { "epoch": 0.0955349192214731, "ewc_loss": 0.0032806396484375, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.137853145599365, "learning_rate": 3.1793132683340396e-07, "loss": 0.5287, "mean_token_accuracy": 0.8320337533950806, "num_tokens": 28542336.0, "step": 751 }, { "epoch": 0.0956621295000636, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 5.1035261154174805, "learning_rate": 3.1835523526918186e-07, "loss": 0.4787, "mean_token_accuracy": 0.8446619510650635, "num_tokens": 28578587.0, "step": 752 }, { "epoch": 0.09578933977865411, "ewc_loss": 0.0032501220703125, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 5.115482807159424, "learning_rate": 3.1877914370495975e-07, "loss": 0.5586, "mean_token_accuracy": 0.8228185176849365, "num_tokens": 28617505.0, "step": 753 }, { "epoch": 0.09591655005724463, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.1495561599731445, "learning_rate": 3.1920305214073755e-07, "loss": 0.5643, "mean_token_accuracy": 0.8226670622825623, "num_tokens": 28658526.0, "step": 754 }, { "epoch": 0.09604376033583513, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.080255031585693, "learning_rate": 3.1962696057651545e-07, "loss": 0.5, "mean_token_accuracy": 0.8401574492454529, "num_tokens": 28698825.0, "step": 755 }, { "epoch": 0.09617097061442564, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.176278591156006, "learning_rate": 3.2005086901229335e-07, "loss": 0.5428, "mean_token_accuracy": 0.8248233795166016, "num_tokens": 28733238.0, "step": 756 }, { "epoch": 0.09629818089301616, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.138284206390381, "learning_rate": 3.2047477744807125e-07, "loss": 0.5438, "mean_token_accuracy": 0.8301187753677368, "num_tokens": 28771552.0, "step": 757 }, { "epoch": 0.09642539117160667, "ewc_loss": 0.0032806396484375, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 5.099484443664551, "learning_rate": 3.2089868588384904e-07, "loss": 0.5743, "mean_token_accuracy": 0.8159473538398743, "num_tokens": 28810827.0, "step": 758 }, { "epoch": 0.09655260145019717, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.091747283935547, "learning_rate": 3.2132259431962694e-07, "loss": 0.4879, "mean_token_accuracy": 0.8445367217063904, "num_tokens": 28849017.0, "step": 759 }, { "epoch": 0.09667981172878769, "ewc_loss": 0.0033111572265625, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 5.185678958892822, "learning_rate": 3.2174650275540484e-07, "loss": 0.538, "mean_token_accuracy": 0.8341125249862671, "num_tokens": 28887659.0, "step": 760 }, { "epoch": 0.0968070220073782, "ewc_loss": 0.0033111572265625, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 5.107750415802002, "learning_rate": 3.2217041119118274e-07, "loss": 0.5043, "mean_token_accuracy": 0.8382952213287354, "num_tokens": 28928983.0, "step": 761 }, { "epoch": 0.0969342322859687, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 5.227722644805908, "learning_rate": 3.2259431962696053e-07, "loss": 0.5371, "mean_token_accuracy": 0.8297967910766602, "num_tokens": 28962195.0, "step": 762 }, { "epoch": 0.09706144256455922, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.150015830993652, "learning_rate": 3.2301822806273843e-07, "loss": 0.5195, "mean_token_accuracy": 0.8324546813964844, "num_tokens": 28997351.0, "step": 763 }, { "epoch": 0.09718865284314973, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 5.134615898132324, "learning_rate": 3.2344213649851633e-07, "loss": 0.546, "mean_token_accuracy": 0.8267743587493896, "num_tokens": 29040988.0, "step": 764 }, { "epoch": 0.09731586312174023, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.085514068603516, "learning_rate": 3.238660449342942e-07, "loss": 0.5086, "mean_token_accuracy": 0.835207998752594, "num_tokens": 29081873.0, "step": 765 }, { "epoch": 0.09744307340033075, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.153860092163086, "learning_rate": 3.24289953370072e-07, "loss": 0.4801, "mean_token_accuracy": 0.8426387906074524, "num_tokens": 29115360.0, "step": 766 }, { "epoch": 0.09757028367892126, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.103602409362793, "learning_rate": 3.247138618058499e-07, "loss": 0.5248, "mean_token_accuracy": 0.8272258043289185, "num_tokens": 29160139.0, "step": 767 }, { "epoch": 0.09769749395751176, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.095025062561035, "learning_rate": 3.251377702416278e-07, "loss": 0.5543, "mean_token_accuracy": 0.8244476318359375, "num_tokens": 29202872.0, "step": 768 }, { "epoch": 0.09782470423610228, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.124780178070068, "learning_rate": 3.255616786774057e-07, "loss": 0.5016, "mean_token_accuracy": 0.8392095565795898, "num_tokens": 29246965.0, "step": 769 }, { "epoch": 0.09795191451469279, "ewc_loss": 0.0034027099609375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 5.06107759475708, "learning_rate": 3.259855871131835e-07, "loss": 0.4413, "mean_token_accuracy": 0.855922281742096, "num_tokens": 29289531.0, "step": 770 }, { "epoch": 0.0980791247932833, "ewc_loss": 0.0033416748046875, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 5.0960469245910645, "learning_rate": 3.264094955489614e-07, "loss": 0.498, "mean_token_accuracy": 0.839372992515564, "num_tokens": 29330484.0, "step": 771 }, { "epoch": 0.0982063350718738, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.147069931030273, "learning_rate": 3.268334039847393e-07, "loss": 0.5247, "mean_token_accuracy": 0.8301959037780762, "num_tokens": 29369595.0, "step": 772 }, { "epoch": 0.09833354535046432, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 5.1762895584106445, "learning_rate": 3.2725731242051715e-07, "loss": 0.5131, "mean_token_accuracy": 0.838508129119873, "num_tokens": 29411856.0, "step": 773 }, { "epoch": 0.09846075562905483, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 5.103264808654785, "learning_rate": 3.27681220856295e-07, "loss": 0.4959, "mean_token_accuracy": 0.8420644998550415, "num_tokens": 29454754.0, "step": 774 }, { "epoch": 0.09858796590764533, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.174259662628174, "learning_rate": 3.281051292920729e-07, "loss": 0.4915, "mean_token_accuracy": 0.8388323783874512, "num_tokens": 29489131.0, "step": 775 }, { "epoch": 0.09871517618623585, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 5.124715328216553, "learning_rate": 3.285290377278508e-07, "loss": 0.546, "mean_token_accuracy": 0.8234477043151855, "num_tokens": 29528292.0, "step": 776 }, { "epoch": 0.09884238646482636, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.169638633728027, "learning_rate": 3.2895294616362864e-07, "loss": 0.4795, "mean_token_accuracy": 0.8463934063911438, "num_tokens": 29564878.0, "step": 777 }, { "epoch": 0.09896959674341686, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 5.126360893249512, "learning_rate": 3.293768545994065e-07, "loss": 0.4988, "mean_token_accuracy": 0.8377853631973267, "num_tokens": 29602826.0, "step": 778 }, { "epoch": 0.09909680702200738, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 5.1591291427612305, "learning_rate": 3.298007630351844e-07, "loss": 0.5668, "mean_token_accuracy": 0.8219143748283386, "num_tokens": 29640323.0, "step": 779 }, { "epoch": 0.09922401730059789, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 5.136446952819824, "learning_rate": 3.302246714709623e-07, "loss": 0.497, "mean_token_accuracy": 0.8392022252082825, "num_tokens": 29683302.0, "step": 780 }, { "epoch": 0.09935122757918839, "ewc_loss": 0.0034027099609375, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 5.170915603637695, "learning_rate": 3.3064857990674013e-07, "loss": 0.5219, "mean_token_accuracy": 0.8357431888580322, "num_tokens": 29721085.0, "step": 781 }, { "epoch": 0.0994784378577789, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 5.234269142150879, "learning_rate": 3.31072488342518e-07, "loss": 0.4741, "mean_token_accuracy": 0.8456604480743408, "num_tokens": 29753810.0, "step": 782 }, { "epoch": 0.09960564813636942, "ewc_loss": 0.00341796875, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 5.127032279968262, "learning_rate": 3.314963967782959e-07, "loss": 0.5016, "mean_token_accuracy": 0.8400708436965942, "num_tokens": 29796488.0, "step": 783 }, { "epoch": 0.09973285841495994, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 5.14953088760376, "learning_rate": 3.319203052140738e-07, "loss": 0.5117, "mean_token_accuracy": 0.8354030251502991, "num_tokens": 29838181.0, "step": 784 }, { "epoch": 0.09986006869355044, "ewc_loss": 0.003387451171875, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 5.376928329467773, "learning_rate": 3.323442136498516e-07, "loss": 0.4934, "mean_token_accuracy": 0.8424557447433472, "num_tokens": 29868893.0, "step": 785 }, { "epoch": 0.09998727897214095, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.277683734893799, "learning_rate": 3.3276812208562947e-07, "loss": 0.5091, "mean_token_accuracy": 0.8349581956863403, "num_tokens": 29903571.0, "step": 786 }, { "epoch": 0.10011448925073146, "ewc_loss": 0.00341796875, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 5.185192584991455, "learning_rate": 3.3319203052140737e-07, "loss": 0.5666, "mean_token_accuracy": 0.8263028860092163, "num_tokens": 29940438.0, "step": 787 }, { "epoch": 0.10024169952932196, "ewc_loss": 0.00341796875, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 5.212259769439697, "learning_rate": 3.336159389571852e-07, "loss": 0.4889, "mean_token_accuracy": 0.8466787934303284, "num_tokens": 29976538.0, "step": 788 }, { "epoch": 0.10036890980791248, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.14049768447876, "learning_rate": 3.340398473929631e-07, "loss": 0.5588, "mean_token_accuracy": 0.822368323802948, "num_tokens": 30019814.0, "step": 789 }, { "epoch": 0.100496120086503, "ewc_loss": 0.0034332275390625, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 5.198901176452637, "learning_rate": 3.3446375582874096e-07, "loss": 0.509, "mean_token_accuracy": 0.8377522230148315, "num_tokens": 30057578.0, "step": 790 }, { "epoch": 0.1006233303650935, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.189035415649414, "learning_rate": 3.3488766426451886e-07, "loss": 0.581, "mean_token_accuracy": 0.8234432339668274, "num_tokens": 30094869.0, "step": 791 }, { "epoch": 0.10075054064368401, "ewc_loss": 0.00341796875, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 5.165081024169922, "learning_rate": 3.353115727002967e-07, "loss": 0.5423, "mean_token_accuracy": 0.8278449773788452, "num_tokens": 30136446.0, "step": 792 }, { "epoch": 0.10087775092227452, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 5.120593070983887, "learning_rate": 3.357354811360746e-07, "loss": 0.5359, "mean_token_accuracy": 0.8247353434562683, "num_tokens": 30179226.0, "step": 793 }, { "epoch": 0.10100496120086502, "ewc_loss": 0.0034332275390625, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 5.192400932312012, "learning_rate": 3.3615938957185245e-07, "loss": 0.5267, "mean_token_accuracy": 0.8316382765769958, "num_tokens": 30213890.0, "step": 794 }, { "epoch": 0.10113217147945554, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 5.276178359985352, "learning_rate": 3.3658329800763035e-07, "loss": 0.5763, "mean_token_accuracy": 0.8138907551765442, "num_tokens": 30245691.0, "step": 795 }, { "epoch": 0.10125938175804605, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 5.1827392578125, "learning_rate": 3.370072064434082e-07, "loss": 0.5318, "mean_token_accuracy": 0.8228020668029785, "num_tokens": 30281738.0, "step": 796 }, { "epoch": 0.10138659203663657, "ewc_loss": 0.00347900390625, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 5.1536078453063965, "learning_rate": 3.374311148791861e-07, "loss": 0.4966, "mean_token_accuracy": 0.8381803035736084, "num_tokens": 30320707.0, "step": 797 }, { "epoch": 0.10151380231522707, "ewc_loss": 0.00347900390625, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 5.381799221038818, "learning_rate": 3.3785502331496394e-07, "loss": 0.5049, "mean_token_accuracy": 0.8354204893112183, "num_tokens": 30351379.0, "step": 798 }, { "epoch": 0.10164101259381758, "ewc_loss": 0.0035400390625, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.342132568359375, "learning_rate": 3.3827893175074184e-07, "loss": 0.5287, "mean_token_accuracy": 0.8291662335395813, "num_tokens": 30382395.0, "step": 799 }, { "epoch": 0.1017682228724081, "ewc_loss": 0.0035247802734375, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.8312206268310547e-06, "grad_norm": 5.237147808074951, "learning_rate": 3.387028401865197e-07, "loss": 0.5347, "mean_token_accuracy": 0.8333403468132019, "num_tokens": 30419758.0, "step": 800 }, { "epoch": 0.1018954331509986, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.195949077606201, "learning_rate": 3.391267486222976e-07, "loss": 0.5187, "mean_token_accuracy": 0.8315474390983582, "num_tokens": 30458115.0, "step": 801 }, { "epoch": 0.10202264342958911, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 5.363593101501465, "learning_rate": 3.3955065705807543e-07, "loss": 0.553, "mean_token_accuracy": 0.8212552070617676, "num_tokens": 30489679.0, "step": 802 }, { "epoch": 0.10214985370817962, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.24785041809082, "learning_rate": 3.3997456549385333e-07, "loss": 0.4898, "mean_token_accuracy": 0.8434954881668091, "num_tokens": 30524874.0, "step": 803 }, { "epoch": 0.10227706398677013, "ewc_loss": 0.0035400390625, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.8461217880249023e-06, "grad_norm": 5.266778945922852, "learning_rate": 3.403984739296312e-07, "loss": 0.563, "mean_token_accuracy": 0.8219352960586548, "num_tokens": 30558354.0, "step": 804 }, { "epoch": 0.10240427426536064, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.225597858428955, "learning_rate": 3.408223823654091e-07, "loss": 0.494, "mean_token_accuracy": 0.8405125141143799, "num_tokens": 30594827.0, "step": 805 }, { "epoch": 0.10253148454395115, "ewc_loss": 0.0035552978515625, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.286140441894531, "learning_rate": 3.412462908011869e-07, "loss": 0.5353, "mean_token_accuracy": 0.8349188566207886, "num_tokens": 30635259.0, "step": 806 }, { "epoch": 0.10265869482254165, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.208769798278809, "learning_rate": 3.4167019923696477e-07, "loss": 0.5278, "mean_token_accuracy": 0.8360694646835327, "num_tokens": 30681850.0, "step": 807 }, { "epoch": 0.10278590510113217, "ewc_loss": 0.0035552978515625, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 5.255061149597168, "learning_rate": 3.4209410767274267e-07, "loss": 0.4945, "mean_token_accuracy": 0.8447257280349731, "num_tokens": 30721302.0, "step": 808 }, { "epoch": 0.10291311537972268, "ewc_loss": 0.0035858154296875, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.3210272789001465, "learning_rate": 3.4251801610852057e-07, "loss": 0.5696, "mean_token_accuracy": 0.8163323402404785, "num_tokens": 30761975.0, "step": 809 }, { "epoch": 0.1030403256583132, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.316085338592529, "learning_rate": 3.429419245442984e-07, "loss": 0.5217, "mean_token_accuracy": 0.8319028615951538, "num_tokens": 30797754.0, "step": 810 }, { "epoch": 0.1031675359369037, "ewc_loss": 0.0035858154296875, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.243374824523926, "learning_rate": 3.4336583298007626e-07, "loss": 0.4962, "mean_token_accuracy": 0.8416273593902588, "num_tokens": 30834183.0, "step": 811 }, { "epoch": 0.10329474621549421, "ewc_loss": 0.0035552978515625, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.158416271209717, "learning_rate": 3.4378974141585416e-07, "loss": 0.5231, "mean_token_accuracy": 0.8334677219390869, "num_tokens": 30874160.0, "step": 812 }, { "epoch": 0.10342195649408473, "ewc_loss": 0.0035552978515625, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 5.218460559844971, "learning_rate": 3.4421364985163206e-07, "loss": 0.4612, "mean_token_accuracy": 0.8498212695121765, "num_tokens": 30916827.0, "step": 813 }, { "epoch": 0.10354916677267523, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.205958843231201, "learning_rate": 3.446375582874099e-07, "loss": 0.5326, "mean_token_accuracy": 0.8314529657363892, "num_tokens": 30960063.0, "step": 814 }, { "epoch": 0.10367637705126574, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 5.274761199951172, "learning_rate": 3.4506146672318775e-07, "loss": 0.4799, "mean_token_accuracy": 0.8429941534996033, "num_tokens": 30995172.0, "step": 815 }, { "epoch": 0.10380358732985626, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.291352272033691, "learning_rate": 3.4548537515896565e-07, "loss": 0.5395, "mean_token_accuracy": 0.8281568288803101, "num_tokens": 31035089.0, "step": 816 }, { "epoch": 0.10393079760844676, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.237557888031006, "learning_rate": 3.4590928359474355e-07, "loss": 0.5221, "mean_token_accuracy": 0.8321337699890137, "num_tokens": 31074836.0, "step": 817 }, { "epoch": 0.10405800788703727, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 5.247933864593506, "learning_rate": 3.463331920305214e-07, "loss": 0.5411, "mean_token_accuracy": 0.8264926671981812, "num_tokens": 31114700.0, "step": 818 }, { "epoch": 0.10418521816562779, "ewc_loss": 0.0035858154296875, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 5.3592352867126465, "learning_rate": 3.4675710046629924e-07, "loss": 0.4892, "mean_token_accuracy": 0.842069149017334, "num_tokens": 31148258.0, "step": 819 }, { "epoch": 0.1043124284442183, "ewc_loss": 0.0036163330078125, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.279374599456787, "learning_rate": 3.4718100890207714e-07, "loss": 0.5286, "mean_token_accuracy": 0.83183753490448, "num_tokens": 31185720.0, "step": 820 }, { "epoch": 0.1044396387228088, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.364620685577393, "learning_rate": 3.4760491733785504e-07, "loss": 0.5392, "mean_token_accuracy": 0.8250612020492554, "num_tokens": 31220353.0, "step": 821 }, { "epoch": 0.10456684900139931, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 2.950429916381836e-06, "grad_norm": 5.321588516235352, "learning_rate": 3.480288257736329e-07, "loss": 0.5167, "mean_token_accuracy": 0.831635594367981, "num_tokens": 31256477.0, "step": 822 }, { "epoch": 0.10469405927998983, "ewc_loss": 0.0036163330078125, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.271932125091553, "learning_rate": 3.4845273420941073e-07, "loss": 0.5154, "mean_token_accuracy": 0.8361260890960693, "num_tokens": 31294365.0, "step": 823 }, { "epoch": 0.10482126955858033, "ewc_loss": 0.0036163330078125, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.409441947937012, "learning_rate": 3.4887664264518863e-07, "loss": 0.5109, "mean_token_accuracy": 0.8327886462211609, "num_tokens": 31329357.0, "step": 824 }, { "epoch": 0.10494847983717084, "ewc_loss": 0.0036773681640625, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.244519233703613, "learning_rate": 3.4930055108096653e-07, "loss": 0.5157, "mean_token_accuracy": 0.8348777294158936, "num_tokens": 31372180.0, "step": 825 }, { "epoch": 0.10507569011576136, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.238104343414307, "learning_rate": 3.497244595167443e-07, "loss": 0.499, "mean_token_accuracy": 0.8387582302093506, "num_tokens": 31414866.0, "step": 826 }, { "epoch": 0.10520290039435186, "ewc_loss": 0.00360107421875, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.905726432800293e-06, "grad_norm": 5.386203765869141, "learning_rate": 3.501483679525222e-07, "loss": 0.5675, "mean_token_accuracy": 0.8184089660644531, "num_tokens": 31450583.0, "step": 827 }, { "epoch": 0.10533011067294237, "ewc_loss": 0.0036773681640625, "ewc_loss_diag": 7.033348083496094e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.3021955490112305, "learning_rate": 3.505722763883001e-07, "loss": 0.4993, "mean_token_accuracy": 0.8375866413116455, "num_tokens": 31490371.0, "step": 828 }, { "epoch": 0.10545732095153289, "ewc_loss": 0.0036163330078125, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 5.344424247741699, "learning_rate": 3.50996184824078e-07, "loss": 0.5409, "mean_token_accuracy": 0.8257996439933777, "num_tokens": 31525951.0, "step": 829 }, { "epoch": 0.10558453123012339, "ewc_loss": 0.003662109375, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.2509331703186035, "learning_rate": 3.514200932598558e-07, "loss": 0.5808, "mean_token_accuracy": 0.8161408305168152, "num_tokens": 31570109.0, "step": 830 }, { "epoch": 0.1057117415087139, "ewc_loss": 0.003662109375, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.37689733505249, "learning_rate": 3.518440016956337e-07, "loss": 0.5232, "mean_token_accuracy": 0.8328319787979126, "num_tokens": 31610569.0, "step": 831 }, { "epoch": 0.10583895178730442, "ewc_loss": 0.0036773681640625, "ewc_loss_diag": 7.033348083496094e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.2816691398620605, "learning_rate": 3.522679101314116e-07, "loss": 0.5347, "mean_token_accuracy": 0.8322160840034485, "num_tokens": 31653585.0, "step": 832 }, { "epoch": 0.10596616206589493, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 2.950429916381836e-06, "grad_norm": 5.270092010498047, "learning_rate": 3.526918185671895e-07, "loss": 0.5444, "mean_token_accuracy": 0.8254210948944092, "num_tokens": 31696353.0, "step": 833 }, { "epoch": 0.10609337234448543, "ewc_loss": 0.003662109375, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.301293849945068, "learning_rate": 3.531157270029673e-07, "loss": 0.4756, "mean_token_accuracy": 0.847080409526825, "num_tokens": 31730620.0, "step": 834 }, { "epoch": 0.10622058262307595, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.31436824798584, "learning_rate": 3.535396354387452e-07, "loss": 0.4916, "mean_token_accuracy": 0.84022057056427, "num_tokens": 31770538.0, "step": 835 }, { "epoch": 0.10634779290166646, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.350349426269531, "learning_rate": 3.539635438745231e-07, "loss": 0.5339, "mean_token_accuracy": 0.8289308547973633, "num_tokens": 31808687.0, "step": 836 }, { "epoch": 0.10647500318025696, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 7.152557373046875e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.252893447875977, "learning_rate": 3.54387452310301e-07, "loss": 0.5127, "mean_token_accuracy": 0.8361152410507202, "num_tokens": 31852310.0, "step": 837 }, { "epoch": 0.10660221345884748, "ewc_loss": 0.0036773681640625, "ewc_loss_diag": 7.152557373046875e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 5.301070690155029, "learning_rate": 3.548113607460788e-07, "loss": 0.4817, "mean_token_accuracy": 0.8474609851837158, "num_tokens": 31887897.0, "step": 838 }, { "epoch": 0.10672942373743799, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.152557373046875e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.335393905639648, "learning_rate": 3.552352691818567e-07, "loss": 0.5069, "mean_token_accuracy": 0.838659942150116, "num_tokens": 31926159.0, "step": 839 }, { "epoch": 0.10685663401602849, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.18235969543457e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.294885635375977, "learning_rate": 3.556591776176346e-07, "loss": 0.4812, "mean_token_accuracy": 0.8458354473114014, "num_tokens": 31968432.0, "step": 840 }, { "epoch": 0.106983844294619, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 7.152557373046875e-06, "ewc_loss_parallel": 2.9802322387695312e-06, "grad_norm": 5.411712169647217, "learning_rate": 3.560830860534125e-07, "loss": 0.5148, "mean_token_accuracy": 0.8375118374824524, "num_tokens": 32000949.0, "step": 841 }, { "epoch": 0.10711105457320952, "ewc_loss": 0.003753662109375, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.4323859214782715, "learning_rate": 3.565069944891903e-07, "loss": 0.5569, "mean_token_accuracy": 0.8217464089393616, "num_tokens": 32033862.0, "step": 842 }, { "epoch": 0.10723826485180002, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 5.372573375701904, "learning_rate": 3.569309029249682e-07, "loss": 0.5194, "mean_token_accuracy": 0.8355488777160645, "num_tokens": 32068623.0, "step": 843 }, { "epoch": 0.10736547513039053, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.271766662597656e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.407020092010498, "learning_rate": 3.573548113607461e-07, "loss": 0.5065, "mean_token_accuracy": 0.8389368653297424, "num_tokens": 32106658.0, "step": 844 }, { "epoch": 0.10749268540898105, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.334863662719727, "learning_rate": 3.577787197965239e-07, "loss": 0.5482, "mean_token_accuracy": 0.824044942855835, "num_tokens": 32142031.0, "step": 845 }, { "epoch": 0.10761989568757156, "ewc_loss": 0.00372314453125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 5.245335578918457, "learning_rate": 3.5820262823230177e-07, "loss": 0.4803, "mean_token_accuracy": 0.8491154313087463, "num_tokens": 32182686.0, "step": 846 }, { "epoch": 0.10774710596616206, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.418031215667725, "learning_rate": 3.5862653666807967e-07, "loss": 0.5753, "mean_token_accuracy": 0.8165870904922485, "num_tokens": 32218027.0, "step": 847 }, { "epoch": 0.10787431624475258, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 5.329488277435303, "learning_rate": 3.5905044510385757e-07, "loss": 0.531, "mean_token_accuracy": 0.8334475755691528, "num_tokens": 32257854.0, "step": 848 }, { "epoch": 0.10800152652334309, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 7.361173629760742e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 5.319369792938232, "learning_rate": 3.594743535396354e-07, "loss": 0.5139, "mean_token_accuracy": 0.8356735706329346, "num_tokens": 32295019.0, "step": 849 }, { "epoch": 0.10812873680193359, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.348956108093262, "learning_rate": 3.5989826197541326e-07, "loss": 0.5631, "mean_token_accuracy": 0.8202724456787109, "num_tokens": 32335652.0, "step": 850 }, { "epoch": 0.1082559470805241, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.039836883544922e-06, "grad_norm": 5.298275947570801, "learning_rate": 3.6032217041119116e-07, "loss": 0.5108, "mean_token_accuracy": 0.8327062129974365, "num_tokens": 32377249.0, "step": 851 }, { "epoch": 0.10838315735911462, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 5.28760290145874, "learning_rate": 3.6074607884696906e-07, "loss": 0.4816, "mean_token_accuracy": 0.8470339775085449, "num_tokens": 32422467.0, "step": 852 }, { "epoch": 0.10851036763770512, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 5.297539710998535, "learning_rate": 3.611699872827469e-07, "loss": 0.4721, "mean_token_accuracy": 0.8456903100013733, "num_tokens": 32462120.0, "step": 853 }, { "epoch": 0.10863757791629564, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.420778274536133e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.436492443084717, "learning_rate": 3.6159389571852475e-07, "loss": 0.5923, "mean_token_accuracy": 0.8128188252449036, "num_tokens": 32498484.0, "step": 854 }, { "epoch": 0.10876478819488615, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.334477424621582, "learning_rate": 3.6201780415430265e-07, "loss": 0.5282, "mean_token_accuracy": 0.8316554427146912, "num_tokens": 32541536.0, "step": 855 }, { "epoch": 0.10889199847347665, "ewc_loss": 0.0037841796875, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 5.406450271606445, "learning_rate": 3.6244171259008055e-07, "loss": 0.5513, "mean_token_accuracy": 0.8282339572906494, "num_tokens": 32579750.0, "step": 856 }, { "epoch": 0.10901920875206716, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.380082130432129, "learning_rate": 3.628656210258584e-07, "loss": 0.5591, "mean_token_accuracy": 0.8204159140586853, "num_tokens": 32617518.0, "step": 857 }, { "epoch": 0.10914641903065768, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.069639205932617e-06, "grad_norm": 5.342731952667236, "learning_rate": 3.6328952946163624e-07, "loss": 0.5008, "mean_token_accuracy": 0.839173436164856, "num_tokens": 32653931.0, "step": 858 }, { "epoch": 0.1092736293092482, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.363893032073975, "learning_rate": 3.6371343789741414e-07, "loss": 0.4889, "mean_token_accuracy": 0.8425836563110352, "num_tokens": 32693797.0, "step": 859 }, { "epoch": 0.1094008395878387, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.4007792472839355, "learning_rate": 3.6413734633319204e-07, "loss": 0.5368, "mean_token_accuracy": 0.829888105392456, "num_tokens": 32732432.0, "step": 860 }, { "epoch": 0.10952804986642921, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.389284133911133, "learning_rate": 3.645612547689699e-07, "loss": 0.4902, "mean_token_accuracy": 0.838280439376831, "num_tokens": 32769225.0, "step": 861 }, { "epoch": 0.10965526014501972, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.320274829864502, "learning_rate": 3.6498516320474773e-07, "loss": 0.4834, "mean_token_accuracy": 0.8447521328926086, "num_tokens": 32812027.0, "step": 862 }, { "epoch": 0.10978247042361022, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.356029033660889, "learning_rate": 3.6540907164052563e-07, "loss": 0.4629, "mean_token_accuracy": 0.8516244888305664, "num_tokens": 32849334.0, "step": 863 }, { "epoch": 0.10990968070220074, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.447265148162842, "learning_rate": 3.658329800763035e-07, "loss": 0.5565, "mean_token_accuracy": 0.8210693597793579, "num_tokens": 32888422.0, "step": 864 }, { "epoch": 0.11003689098079125, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.437808990478516, "learning_rate": 3.662568885120814e-07, "loss": 0.516, "mean_token_accuracy": 0.833618700504303, "num_tokens": 32922672.0, "step": 865 }, { "epoch": 0.11016410125938175, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.453664302825928, "learning_rate": 3.666807969478592e-07, "loss": 0.546, "mean_token_accuracy": 0.8228939771652222, "num_tokens": 32961694.0, "step": 866 }, { "epoch": 0.11029131153797227, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.377868175506592, "learning_rate": 3.671047053836371e-07, "loss": 0.5343, "mean_token_accuracy": 0.8269752264022827, "num_tokens": 32998595.0, "step": 867 }, { "epoch": 0.11041852181656278, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.338855266571045, "learning_rate": 3.6752861381941497e-07, "loss": 0.4949, "mean_token_accuracy": 0.8397282361984253, "num_tokens": 33035884.0, "step": 868 }, { "epoch": 0.11054573209515328, "ewc_loss": 0.0038299560546875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 5.4486308097839355, "learning_rate": 3.6795252225519287e-07, "loss": 0.5562, "mean_token_accuracy": 0.8252303600311279, "num_tokens": 33073335.0, "step": 869 }, { "epoch": 0.1106729423737438, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.366863250732422, "learning_rate": 3.6837643069097077e-07, "loss": 0.4561, "mean_token_accuracy": 0.8514032363891602, "num_tokens": 33111443.0, "step": 870 }, { "epoch": 0.11080015265233431, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.352163791656494, "learning_rate": 3.688003391267486e-07, "loss": 0.516, "mean_token_accuracy": 0.8378231525421143, "num_tokens": 33156362.0, "step": 871 }, { "epoch": 0.11092736293092482, "ewc_loss": 0.00384521484375, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 5.326137542724609, "learning_rate": 3.6922424756252646e-07, "loss": 0.485, "mean_token_accuracy": 0.8433510065078735, "num_tokens": 33204968.0, "step": 872 }, { "epoch": 0.11105457320951533, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.435885429382324, "learning_rate": 3.6964815599830436e-07, "loss": 0.532, "mean_token_accuracy": 0.8274696469306946, "num_tokens": 33239265.0, "step": 873 }, { "epoch": 0.11118178348810584, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.465987682342529, "learning_rate": 3.7007206443408226e-07, "loss": 0.4654, "mean_token_accuracy": 0.8509633541107178, "num_tokens": 33275897.0, "step": 874 }, { "epoch": 0.11130899376669635, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.424160480499268, "learning_rate": 3.704959728698601e-07, "loss": 0.588, "mean_token_accuracy": 0.8143208026885986, "num_tokens": 33313966.0, "step": 875 }, { "epoch": 0.11143620404528685, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.419533729553223, "learning_rate": 3.7091988130563795e-07, "loss": 0.5337, "mean_token_accuracy": 0.8300058841705322, "num_tokens": 33352909.0, "step": 876 }, { "epoch": 0.11156341432387737, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.34049129486084, "learning_rate": 3.7134378974141585e-07, "loss": 0.5161, "mean_token_accuracy": 0.8371677398681641, "num_tokens": 33400120.0, "step": 877 }, { "epoch": 0.11169062460246788, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.4542412757873535, "learning_rate": 3.7176769817719375e-07, "loss": 0.5432, "mean_token_accuracy": 0.8270336389541626, "num_tokens": 33436703.0, "step": 878 }, { "epoch": 0.11181783488105838, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.408092021942139, "learning_rate": 3.7219160661297154e-07, "loss": 0.4526, "mean_token_accuracy": 0.8553591966629028, "num_tokens": 33473515.0, "step": 879 }, { "epoch": 0.1119450451596489, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.4338884353637695, "learning_rate": 3.7261551504874944e-07, "loss": 0.5148, "mean_token_accuracy": 0.8369611501693726, "num_tokens": 33507659.0, "step": 880 }, { "epoch": 0.11207225543823941, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.410878658294678, "learning_rate": 3.7303942348452734e-07, "loss": 0.4607, "mean_token_accuracy": 0.8496057987213135, "num_tokens": 33542430.0, "step": 881 }, { "epoch": 0.11219946571682991, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.367368698120117, "learning_rate": 3.7346333192030524e-07, "loss": 0.4679, "mean_token_accuracy": 0.8525028824806213, "num_tokens": 33582579.0, "step": 882 }, { "epoch": 0.11232667599542043, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.381000995635986, "learning_rate": 3.7388724035608303e-07, "loss": 0.4916, "mean_token_accuracy": 0.845454752445221, "num_tokens": 33625928.0, "step": 883 }, { "epoch": 0.11245388627401094, "ewc_loss": 0.0038909912109375, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.450543403625488, "learning_rate": 3.7431114879186093e-07, "loss": 0.5328, "mean_token_accuracy": 0.8313140273094177, "num_tokens": 33661325.0, "step": 884 }, { "epoch": 0.11258109655260146, "ewc_loss": 0.0038909912109375, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.43666934967041, "learning_rate": 3.7473505722763883e-07, "loss": 0.5546, "mean_token_accuracy": 0.8225964307785034, "num_tokens": 33696472.0, "step": 885 }, { "epoch": 0.11270830683119196, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 7.599592208862305e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 5.44244384765625, "learning_rate": 3.7515896566341673e-07, "loss": 0.5092, "mean_token_accuracy": 0.8341153264045715, "num_tokens": 33732637.0, "step": 886 }, { "epoch": 0.11283551710978247, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.400625228881836, "learning_rate": 3.755828740991945e-07, "loss": 0.5332, "mean_token_accuracy": 0.8297773003578186, "num_tokens": 33775106.0, "step": 887 }, { "epoch": 0.11296272738837299, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.429153919219971, "learning_rate": 3.760067825349724e-07, "loss": 0.4959, "mean_token_accuracy": 0.8426986932754517, "num_tokens": 33811297.0, "step": 888 }, { "epoch": 0.11308993766696349, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.3838372230529785, "learning_rate": 3.764306909707503e-07, "loss": 0.4579, "mean_token_accuracy": 0.8508948683738708, "num_tokens": 33854038.0, "step": 889 }, { "epoch": 0.113217147945554, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.493419170379639, "learning_rate": 3.768545994065282e-07, "loss": 0.5144, "mean_token_accuracy": 0.8346776962280273, "num_tokens": 33884929.0, "step": 890 }, { "epoch": 0.11334435822414451, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 5.446438312530518, "learning_rate": 3.77278507842306e-07, "loss": 0.5226, "mean_token_accuracy": 0.8339977264404297, "num_tokens": 33926065.0, "step": 891 }, { "epoch": 0.11347156850273502, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 5.375667095184326, "learning_rate": 3.777024162780839e-07, "loss": 0.4731, "mean_token_accuracy": 0.8444575071334839, "num_tokens": 33965557.0, "step": 892 }, { "epoch": 0.11359877878132553, "ewc_loss": 0.00390625, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 5.447597980499268, "learning_rate": 3.781263247138618e-07, "loss": 0.4819, "mean_token_accuracy": 0.8436844348907471, "num_tokens": 34004132.0, "step": 893 }, { "epoch": 0.11372598905991604, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.68899917602539e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 5.474415302276611, "learning_rate": 3.785502331496397e-07, "loss": 0.5094, "mean_token_accuracy": 0.829362154006958, "num_tokens": 34039781.0, "step": 894 }, { "epoch": 0.11385319933850654, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.1888484954833984e-06, "grad_norm": 5.4251580238342285, "learning_rate": 3.789741415854175e-07, "loss": 0.5354, "mean_token_accuracy": 0.8273510932922363, "num_tokens": 34080487.0, "step": 895 }, { "epoch": 0.11398040961709706, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 5.426713943481445, "learning_rate": 3.793980500211954e-07, "loss": 0.5126, "mean_token_accuracy": 0.8349810242652893, "num_tokens": 34119913.0, "step": 896 }, { "epoch": 0.11410761989568757, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 5.421437740325928, "learning_rate": 3.798219584569733e-07, "loss": 0.4469, "mean_token_accuracy": 0.8547780513763428, "num_tokens": 34155811.0, "step": 897 }, { "epoch": 0.11423483017427809, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.527626991271973, "learning_rate": 3.8024586689275115e-07, "loss": 0.5683, "mean_token_accuracy": 0.8231782913208008, "num_tokens": 34195022.0, "step": 898 }, { "epoch": 0.11436204045286859, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.472961902618408, "learning_rate": 3.80669775328529e-07, "loss": 0.5009, "mean_token_accuracy": 0.8354083299636841, "num_tokens": 34229182.0, "step": 899 }, { "epoch": 0.1144892507314591, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 5.412285327911377, "learning_rate": 3.810936837643069e-07, "loss": 0.4622, "mean_token_accuracy": 0.8489753603935242, "num_tokens": 34266931.0, "step": 900 }, { "epoch": 0.11461646101004962, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 3.1888484954833984e-06, "grad_norm": 5.493655681610107, "learning_rate": 3.815175922000848e-07, "loss": 0.4547, "mean_token_accuracy": 0.8494477272033691, "num_tokens": 34301705.0, "step": 901 }, { "epoch": 0.11474367128864012, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 3.2186508178710938e-06, "grad_norm": 5.489182949066162, "learning_rate": 3.8194150063586264e-07, "loss": 0.5372, "mean_token_accuracy": 0.8326870203018188, "num_tokens": 34340975.0, "step": 902 }, { "epoch": 0.11487088156723063, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 3.1888484954833984e-06, "grad_norm": 5.4583048820495605, "learning_rate": 3.823654090716405e-07, "loss": 0.5052, "mean_token_accuracy": 0.8355646133422852, "num_tokens": 34380024.0, "step": 903 }, { "epoch": 0.11499809184582115, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 3.203749656677246e-06, "grad_norm": 5.4563069343566895, "learning_rate": 3.827893175074184e-07, "loss": 0.5266, "mean_token_accuracy": 0.8382928371429443, "num_tokens": 34419654.0, "step": 904 }, { "epoch": 0.11512530212441165, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2335519790649414e-06, "grad_norm": 5.471973419189453, "learning_rate": 3.832132259431963e-07, "loss": 0.4574, "mean_token_accuracy": 0.8535802960395813, "num_tokens": 34453034.0, "step": 905 }, { "epoch": 0.11525251240300216, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.248453140258789e-06, "grad_norm": 5.470483303070068, "learning_rate": 3.8363713437897413e-07, "loss": 0.5624, "mean_token_accuracy": 0.8197624087333679, "num_tokens": 34493694.0, "step": 906 }, { "epoch": 0.11537972268159268, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 5.496582508087158, "learning_rate": 3.8406104281475197e-07, "loss": 0.5549, "mean_token_accuracy": 0.8248672485351562, "num_tokens": 34531636.0, "step": 907 }, { "epoch": 0.11550693296018319, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.248453140258789e-06, "grad_norm": 5.51521110534668, "learning_rate": 3.8448495125052987e-07, "loss": 0.5391, "mean_token_accuracy": 0.8256171941757202, "num_tokens": 34570851.0, "step": 908 }, { "epoch": 0.11563414323877369, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 5.44236946105957, "learning_rate": 3.8490885968630777e-07, "loss": 0.4913, "mean_token_accuracy": 0.8421502113342285, "num_tokens": 34613314.0, "step": 909 }, { "epoch": 0.1157613535173642, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2335519790649414e-06, "grad_norm": 5.464019775390625, "learning_rate": 3.853327681220856e-07, "loss": 0.5067, "mean_token_accuracy": 0.8359480500221252, "num_tokens": 34656701.0, "step": 910 }, { "epoch": 0.11588856379595472, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.546708106994629, "learning_rate": 3.8575667655786346e-07, "loss": 0.4819, "mean_token_accuracy": 0.8456231355667114, "num_tokens": 34688595.0, "step": 911 }, { "epoch": 0.11601577407454522, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.487269878387451, "learning_rate": 3.8618058499364136e-07, "loss": 0.5003, "mean_token_accuracy": 0.8401665687561035, "num_tokens": 34726609.0, "step": 912 }, { "epoch": 0.11614298435313573, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.4842000007629395, "learning_rate": 3.8660449342941926e-07, "loss": 0.5569, "mean_token_accuracy": 0.825507640838623, "num_tokens": 34769213.0, "step": 913 }, { "epoch": 0.11627019463172625, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 5.5382795333862305, "learning_rate": 3.870284018651971e-07, "loss": 0.5298, "mean_token_accuracy": 0.8312748670578003, "num_tokens": 34806722.0, "step": 914 }, { "epoch": 0.11639740491031675, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.457311153411865, "learning_rate": 3.8745231030097495e-07, "loss": 0.4856, "mean_token_accuracy": 0.8451443314552307, "num_tokens": 34846408.0, "step": 915 }, { "epoch": 0.11652461518890726, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.248453140258789e-06, "grad_norm": 5.484299182891846, "learning_rate": 3.8787621873675285e-07, "loss": 0.4725, "mean_token_accuracy": 0.8467631340026855, "num_tokens": 34883791.0, "step": 916 }, { "epoch": 0.11665182546749778, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 5.490225315093994, "learning_rate": 3.883001271725307e-07, "loss": 0.4688, "mean_token_accuracy": 0.8481139540672302, "num_tokens": 34922768.0, "step": 917 }, { "epoch": 0.11677903574608828, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.489367961883545, "learning_rate": 3.887240356083086e-07, "loss": 0.4483, "mean_token_accuracy": 0.8531348705291748, "num_tokens": 34962184.0, "step": 918 }, { "epoch": 0.11690624602467879, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.633130073547363, "learning_rate": 3.8914794404408644e-07, "loss": 0.4929, "mean_token_accuracy": 0.8400155901908875, "num_tokens": 34991455.0, "step": 919 }, { "epoch": 0.1170334563032693, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 5.5873613357543945, "learning_rate": 3.8957185247986434e-07, "loss": 0.5219, "mean_token_accuracy": 0.830788254737854, "num_tokens": 35025044.0, "step": 920 }, { "epoch": 0.11716066658185982, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 5.488969326019287, "learning_rate": 3.899957609156422e-07, "loss": 0.4949, "mean_token_accuracy": 0.8393747806549072, "num_tokens": 35064564.0, "step": 921 }, { "epoch": 0.11728787686045032, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 5.475819110870361, "learning_rate": 3.904196693514201e-07, "loss": 0.5124, "mean_token_accuracy": 0.8368616104125977, "num_tokens": 35109786.0, "step": 922 }, { "epoch": 0.11741508713904084, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.293156623840332e-06, "grad_norm": 5.554932117462158, "learning_rate": 3.9084357778719793e-07, "loss": 0.5754, "mean_token_accuracy": 0.8168274164199829, "num_tokens": 35148351.0, "step": 923 }, { "epoch": 0.11754229741763135, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.5831685066223145, "learning_rate": 3.9126748622297583e-07, "loss": 0.5391, "mean_token_accuracy": 0.8296536207199097, "num_tokens": 35183464.0, "step": 924 }, { "epoch": 0.11766950769622185, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.548746109008789, "learning_rate": 3.916913946587537e-07, "loss": 0.5219, "mean_token_accuracy": 0.8301427364349365, "num_tokens": 35218561.0, "step": 925 }, { "epoch": 0.11779671797481236, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 5.494829177856445, "learning_rate": 3.921153030945316e-07, "loss": 0.5368, "mean_token_accuracy": 0.8271845579147339, "num_tokens": 35257740.0, "step": 926 }, { "epoch": 0.11792392825340288, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.564457893371582, "learning_rate": 3.925392115303094e-07, "loss": 0.4933, "mean_token_accuracy": 0.8409294486045837, "num_tokens": 35291632.0, "step": 927 }, { "epoch": 0.11805113853199338, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 5.576883792877197, "learning_rate": 3.929631199660873e-07, "loss": 0.4974, "mean_token_accuracy": 0.8369197845458984, "num_tokens": 35327686.0, "step": 928 }, { "epoch": 0.1181783488105839, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.4898271560668945, "learning_rate": 3.9338702840186517e-07, "loss": 0.4985, "mean_token_accuracy": 0.8373283743858337, "num_tokens": 35364480.0, "step": 929 }, { "epoch": 0.11830555908917441, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.502030849456787, "learning_rate": 3.9381093683764307e-07, "loss": 0.5041, "mean_token_accuracy": 0.838750958442688, "num_tokens": 35406017.0, "step": 930 }, { "epoch": 0.11843276936776491, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.507542610168457, "learning_rate": 3.942348452734209e-07, "loss": 0.5631, "mean_token_accuracy": 0.8197983503341675, "num_tokens": 35448447.0, "step": 931 }, { "epoch": 0.11855997964635542, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.519814491271973, "learning_rate": 3.946587537091988e-07, "loss": 0.5119, "mean_token_accuracy": 0.8335537910461426, "num_tokens": 35488120.0, "step": 932 }, { "epoch": 0.11868718992494594, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.48429012298584, "learning_rate": 3.9508266214497666e-07, "loss": 0.4862, "mean_token_accuracy": 0.842646598815918, "num_tokens": 35531759.0, "step": 933 }, { "epoch": 0.11881440020353645, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.507186412811279, "learning_rate": 3.9550657058075456e-07, "loss": 0.4699, "mean_token_accuracy": 0.8495609760284424, "num_tokens": 35571476.0, "step": 934 }, { "epoch": 0.11894161048212695, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 5.6037774085998535, "learning_rate": 3.959304790165324e-07, "loss": 0.5111, "mean_token_accuracy": 0.8371542096138, "num_tokens": 35608440.0, "step": 935 }, { "epoch": 0.11906882076071747, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.601193428039551, "learning_rate": 3.9635438745231025e-07, "loss": 0.5063, "mean_token_accuracy": 0.8387998342514038, "num_tokens": 35641861.0, "step": 936 }, { "epoch": 0.11919603103930798, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.576205253601074, "learning_rate": 3.9677829588808815e-07, "loss": 0.5279, "mean_token_accuracy": 0.8300971388816833, "num_tokens": 35684626.0, "step": 937 }, { "epoch": 0.11932324131789848, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.5923590660095215, "learning_rate": 3.9720220432386605e-07, "loss": 0.5342, "mean_token_accuracy": 0.8277679681777954, "num_tokens": 35718125.0, "step": 938 }, { "epoch": 0.119450451596489, "ewc_loss": 0.004150390625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 5.590816497802734, "learning_rate": 3.976261127596439e-07, "loss": 0.4779, "mean_token_accuracy": 0.8477599024772644, "num_tokens": 35752818.0, "step": 939 }, { "epoch": 0.11957766187507951, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.552156448364258, "learning_rate": 3.9805002119542174e-07, "loss": 0.4583, "mean_token_accuracy": 0.8513237833976746, "num_tokens": 35787097.0, "step": 940 }, { "epoch": 0.11970487215367001, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.555205821990967, "learning_rate": 3.9847392963119964e-07, "loss": 0.5552, "mean_token_accuracy": 0.8259297609329224, "num_tokens": 35825887.0, "step": 941 }, { "epoch": 0.11983208243226053, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 5.529861927032471, "learning_rate": 3.9889783806697754e-07, "loss": 0.4838, "mean_token_accuracy": 0.8461236357688904, "num_tokens": 35863592.0, "step": 942 }, { "epoch": 0.11995929271085104, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.427267074584961e-06, "grad_norm": 5.597880840301514, "learning_rate": 3.993217465027554e-07, "loss": 0.5587, "mean_token_accuracy": 0.8254454135894775, "num_tokens": 35904483.0, "step": 943 }, { "epoch": 0.12008650298944154, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.620852470397949, "learning_rate": 3.9974565493853323e-07, "loss": 0.4458, "mean_token_accuracy": 0.8559551239013672, "num_tokens": 35938662.0, "step": 944 }, { "epoch": 0.12021371326803205, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.611592769622803, "learning_rate": 4.0016956337431113e-07, "loss": 0.5097, "mean_token_accuracy": 0.835219144821167, "num_tokens": 35975176.0, "step": 945 }, { "epoch": 0.12034092354662257, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.582108020782471, "learning_rate": 4.0059347181008903e-07, "loss": 0.4961, "mean_token_accuracy": 0.8395135402679443, "num_tokens": 36008893.0, "step": 946 }, { "epoch": 0.12046813382521308, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.516968727111816, "learning_rate": 4.010173802458669e-07, "loss": 0.4555, "mean_token_accuracy": 0.8543197512626648, "num_tokens": 36049882.0, "step": 947 }, { "epoch": 0.12059534410380358, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.581897735595703, "learning_rate": 4.014412886816447e-07, "loss": 0.5686, "mean_token_accuracy": 0.8218618631362915, "num_tokens": 36090615.0, "step": 948 }, { "epoch": 0.1207225543823941, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.583312034606934, "learning_rate": 4.018651971174226e-07, "loss": 0.4991, "mean_token_accuracy": 0.8380920886993408, "num_tokens": 36128159.0, "step": 949 }, { "epoch": 0.12084976466098461, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 5.530418395996094, "learning_rate": 4.022891055532005e-07, "loss": 0.5132, "mean_token_accuracy": 0.8357176780700684, "num_tokens": 36172851.0, "step": 950 }, { "epoch": 0.12097697493957511, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.586582660675049, "learning_rate": 4.0271301398897837e-07, "loss": 0.5635, "mean_token_accuracy": 0.8233546018600464, "num_tokens": 36214282.0, "step": 951 }, { "epoch": 0.12110418521816563, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.427267074584961e-06, "grad_norm": 5.615633010864258, "learning_rate": 4.031369224247562e-07, "loss": 0.4846, "mean_token_accuracy": 0.8395553231239319, "num_tokens": 36248619.0, "step": 952 }, { "epoch": 0.12123139549675614, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 5.570388317108154, "learning_rate": 4.035608308605341e-07, "loss": 0.5211, "mean_token_accuracy": 0.8310612440109253, "num_tokens": 36289070.0, "step": 953 }, { "epoch": 0.12135860577534664, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.546158313751221, "learning_rate": 4.03984739296312e-07, "loss": 0.5206, "mean_token_accuracy": 0.8352367281913757, "num_tokens": 36330883.0, "step": 954 }, { "epoch": 0.12148581605393716, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.553861141204834, "learning_rate": 4.044086477320898e-07, "loss": 0.5005, "mean_token_accuracy": 0.8397461175918579, "num_tokens": 36371507.0, "step": 955 }, { "epoch": 0.12161302633252767, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.6218719482421875, "learning_rate": 4.048325561678677e-07, "loss": 0.5233, "mean_token_accuracy": 0.8313450813293457, "num_tokens": 36407262.0, "step": 956 }, { "epoch": 0.12174023661111817, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.602742671966553, "learning_rate": 4.052564646036456e-07, "loss": 0.5198, "mean_token_accuracy": 0.8341999053955078, "num_tokens": 36448089.0, "step": 957 }, { "epoch": 0.12186744688970869, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.597552299499512, "learning_rate": 4.056803730394235e-07, "loss": 0.5321, "mean_token_accuracy": 0.8272756934165955, "num_tokens": 36484022.0, "step": 958 }, { "epoch": 0.1219946571682992, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 5.651618003845215, "learning_rate": 4.061042814752013e-07, "loss": 0.5082, "mean_token_accuracy": 0.8380404710769653, "num_tokens": 36519385.0, "step": 959 }, { "epoch": 0.12212186744688971, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.7185378074646, "learning_rate": 4.065281899109792e-07, "loss": 0.5122, "mean_token_accuracy": 0.8343939185142517, "num_tokens": 36551943.0, "step": 960 }, { "epoch": 0.12224907772548022, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 5.5705885887146, "learning_rate": 4.069520983467571e-07, "loss": 0.4587, "mean_token_accuracy": 0.8524246215820312, "num_tokens": 36590927.0, "step": 961 }, { "epoch": 0.12237628800407073, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.427267074584961e-06, "grad_norm": 5.638700008392334, "learning_rate": 4.07376006782535e-07, "loss": 0.4904, "mean_token_accuracy": 0.8396057486534119, "num_tokens": 36625199.0, "step": 962 }, { "epoch": 0.12250349828266124, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.656710624694824, "learning_rate": 4.077999152183128e-07, "loss": 0.4533, "mean_token_accuracy": 0.8527933359146118, "num_tokens": 36658981.0, "step": 963 }, { "epoch": 0.12263070856125174, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.555657863616943, "learning_rate": 4.082238236540907e-07, "loss": 0.5137, "mean_token_accuracy": 0.8383691310882568, "num_tokens": 36704214.0, "step": 964 }, { "epoch": 0.12275791883984226, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.625550746917725, "learning_rate": 4.086477320898686e-07, "loss": 0.4831, "mean_token_accuracy": 0.8424136638641357, "num_tokens": 36744782.0, "step": 965 }, { "epoch": 0.12288512911843277, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.615815162658691, "learning_rate": 4.090716405256465e-07, "loss": 0.5181, "mean_token_accuracy": 0.8320188522338867, "num_tokens": 36783645.0, "step": 966 }, { "epoch": 0.12301233939702327, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.6839704513549805, "learning_rate": 4.094955489614243e-07, "loss": 0.4865, "mean_token_accuracy": 0.841447114944458, "num_tokens": 36817539.0, "step": 967 }, { "epoch": 0.12313954967561379, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.631120204925537, "learning_rate": 4.099194573972022e-07, "loss": 0.488, "mean_token_accuracy": 0.8413589000701904, "num_tokens": 36859144.0, "step": 968 }, { "epoch": 0.1232667599542043, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.611396789550781, "learning_rate": 4.1034336583298007e-07, "loss": 0.5308, "mean_token_accuracy": 0.8298420906066895, "num_tokens": 36900833.0, "step": 969 }, { "epoch": 0.1233939702327948, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.620634078979492, "learning_rate": 4.1076727426875797e-07, "loss": 0.5199, "mean_token_accuracy": 0.8325515985488892, "num_tokens": 36940935.0, "step": 970 }, { "epoch": 0.12352118051138532, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.665501594543457, "learning_rate": 4.1119118270453577e-07, "loss": 0.5817, "mean_token_accuracy": 0.813374400138855, "num_tokens": 36979874.0, "step": 971 }, { "epoch": 0.12364839078997583, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4868717193603516e-06, "grad_norm": 5.612422466278076, "learning_rate": 4.1161509114031366e-07, "loss": 0.4634, "mean_token_accuracy": 0.8524149656295776, "num_tokens": 37017917.0, "step": 972 }, { "epoch": 0.12377560106856635, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 5.68890905380249, "learning_rate": 4.1203899957609156e-07, "loss": 0.5279, "mean_token_accuracy": 0.831583559513092, "num_tokens": 37054914.0, "step": 973 }, { "epoch": 0.12390281134715685, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.721287727355957, "learning_rate": 4.124629080118694e-07, "loss": 0.4782, "mean_token_accuracy": 0.8421685099601746, "num_tokens": 37088277.0, "step": 974 }, { "epoch": 0.12403002162574736, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.5017728805541992e-06, "grad_norm": 5.626038074493408, "learning_rate": 4.1288681644764726e-07, "loss": 0.5237, "mean_token_accuracy": 0.8309118151664734, "num_tokens": 37126963.0, "step": 975 }, { "epoch": 0.12415723190433788, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 5.663862705230713, "learning_rate": 4.1331072488342515e-07, "loss": 0.5369, "mean_token_accuracy": 0.828600287437439, "num_tokens": 37161948.0, "step": 976 }, { "epoch": 0.12428444218292838, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.637840270996094, "learning_rate": 4.1373463331920305e-07, "loss": 0.4518, "mean_token_accuracy": 0.853446900844574, "num_tokens": 37198552.0, "step": 977 }, { "epoch": 0.12441165246151889, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.668240547180176, "learning_rate": 4.141585417549809e-07, "loss": 0.5219, "mean_token_accuracy": 0.8308358192443848, "num_tokens": 37235990.0, "step": 978 }, { "epoch": 0.1245388627401094, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.699736595153809, "learning_rate": 4.1458245019075875e-07, "loss": 0.4814, "mean_token_accuracy": 0.8428124189376831, "num_tokens": 37269664.0, "step": 979 }, { "epoch": 0.1246660730186999, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 5.657586097717285, "learning_rate": 4.1500635862653664e-07, "loss": 0.5127, "mean_token_accuracy": 0.8363226652145386, "num_tokens": 37308647.0, "step": 980 }, { "epoch": 0.12479328329729042, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 5.6571478843688965, "learning_rate": 4.1543026706231454e-07, "loss": 0.4621, "mean_token_accuracy": 0.8510129451751709, "num_tokens": 37345953.0, "step": 981 }, { "epoch": 0.12492049357588093, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.639881610870361, "learning_rate": 4.158541754980924e-07, "loss": 0.4581, "mean_token_accuracy": 0.8538178205490112, "num_tokens": 37382735.0, "step": 982 }, { "epoch": 0.12504770385447145, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.546476364135742e-06, "grad_norm": 5.714675426483154, "learning_rate": 4.1627808393387024e-07, "loss": 0.5246, "mean_token_accuracy": 0.8285894393920898, "num_tokens": 37419605.0, "step": 983 }, { "epoch": 0.12517491413306195, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.591179847717285e-06, "grad_norm": 5.730739116668701, "learning_rate": 4.1670199236964813e-07, "loss": 0.5171, "mean_token_accuracy": 0.8295301198959351, "num_tokens": 37455395.0, "step": 984 }, { "epoch": 0.12530212441165245, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.5315752029418945e-06, "grad_norm": 5.655876159667969, "learning_rate": 4.1712590080542603e-07, "loss": 0.4701, "mean_token_accuracy": 0.8478415012359619, "num_tokens": 37492208.0, "step": 985 }, { "epoch": 0.12542933469024298, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 5.624608516693115, "learning_rate": 4.175498092412039e-07, "loss": 0.5153, "mean_token_accuracy": 0.836682140827179, "num_tokens": 37533675.0, "step": 986 }, { "epoch": 0.12555654496883348, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.679640293121338, "learning_rate": 4.179737176769817e-07, "loss": 0.4798, "mean_token_accuracy": 0.8457902073860168, "num_tokens": 37572792.0, "step": 987 }, { "epoch": 0.12568375524742398, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.6795973777771, "learning_rate": 4.183976261127596e-07, "loss": 0.4765, "mean_token_accuracy": 0.8448293209075928, "num_tokens": 37611481.0, "step": 988 }, { "epoch": 0.1258109655260145, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.713130474090576, "learning_rate": 4.1882153454853747e-07, "loss": 0.5547, "mean_token_accuracy": 0.8218252658843994, "num_tokens": 37648253.0, "step": 989 }, { "epoch": 0.125938175804605, "ewc_loss": 0.00439453125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 5.655181407928467, "learning_rate": 4.1924544298431537e-07, "loss": 0.4885, "mean_token_accuracy": 0.8433519601821899, "num_tokens": 37687086.0, "step": 990 }, { "epoch": 0.12606538608319554, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.591179847717285e-06, "grad_norm": 5.612891674041748, "learning_rate": 4.196693514200932e-07, "loss": 0.4826, "mean_token_accuracy": 0.8439754247665405, "num_tokens": 37730928.0, "step": 991 }, { "epoch": 0.12619259636178604, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.660103797912598, "learning_rate": 4.200932598558711e-07, "loss": 0.5069, "mean_token_accuracy": 0.8343707323074341, "num_tokens": 37771857.0, "step": 992 }, { "epoch": 0.12631980664037654, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.6507844924926758e-06, "grad_norm": 5.635402679443359, "learning_rate": 4.2051716829164896e-07, "loss": 0.5099, "mean_token_accuracy": 0.8387486934661865, "num_tokens": 37817517.0, "step": 993 }, { "epoch": 0.12644701691896706, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.68532657623291, "learning_rate": 4.2094107672742686e-07, "loss": 0.5526, "mean_token_accuracy": 0.828134298324585, "num_tokens": 37855891.0, "step": 994 }, { "epoch": 0.12657422719755757, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.635883331298828e-06, "grad_norm": 5.726948261260986, "learning_rate": 4.2136498516320476e-07, "loss": 0.4726, "mean_token_accuracy": 0.8488678932189941, "num_tokens": 37888719.0, "step": 995 }, { "epoch": 0.12670143747614807, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.619077205657959, "learning_rate": 4.217888935989826e-07, "loss": 0.5243, "mean_token_accuracy": 0.8313552737236023, "num_tokens": 37932968.0, "step": 996 }, { "epoch": 0.1268286477547386, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 5.652359485626221, "learning_rate": 4.2221280203476045e-07, "loss": 0.5164, "mean_token_accuracy": 0.8320851922035217, "num_tokens": 37974297.0, "step": 997 }, { "epoch": 0.1269558580333291, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.660310745239258, "learning_rate": 4.2263671047053835e-07, "loss": 0.5326, "mean_token_accuracy": 0.8298302292823792, "num_tokens": 38020294.0, "step": 998 }, { "epoch": 0.1270830683119196, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.6209821701049805e-06, "grad_norm": 5.6441216468811035, "learning_rate": 4.2306061890631625e-07, "loss": 0.5124, "mean_token_accuracy": 0.8359287977218628, "num_tokens": 38063322.0, "step": 999 }, { "epoch": 0.12721027859051012, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 8.344650268554688e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.726240634918213, "learning_rate": 4.234845273420941e-07, "loss": 0.6006, "mean_token_accuracy": 0.8170943856239319, "num_tokens": 38101052.0, "step": 1000 }, { "epoch": 0.12733748886910062, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.668453216552734, "learning_rate": 4.2390843577787194e-07, "loss": 0.5206, "mean_token_accuracy": 0.8322681188583374, "num_tokens": 38145472.0, "step": 1001 }, { "epoch": 0.12746469914769112, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.674837589263916, "learning_rate": 4.2433234421364984e-07, "loss": 0.4979, "mean_token_accuracy": 0.840343713760376, "num_tokens": 38182662.0, "step": 1002 }, { "epoch": 0.12759190942628165, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 5.720059871673584, "learning_rate": 4.2475625264942774e-07, "loss": 0.4839, "mean_token_accuracy": 0.8456566333770752, "num_tokens": 38224735.0, "step": 1003 }, { "epoch": 0.12771911970487215, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 5.858996391296387, "learning_rate": 4.251801610852056e-07, "loss": 0.5272, "mean_token_accuracy": 0.8303276300430298, "num_tokens": 38255847.0, "step": 1004 }, { "epoch": 0.12784632998346265, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.714203357696533, "learning_rate": 4.2560406952098343e-07, "loss": 0.5403, "mean_token_accuracy": 0.8293672800064087, "num_tokens": 38296921.0, "step": 1005 }, { "epoch": 0.12797354026205318, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.6209821701049805e-06, "grad_norm": 5.686882495880127, "learning_rate": 4.2602797795676133e-07, "loss": 0.5506, "mean_token_accuracy": 0.8250877261161804, "num_tokens": 38336408.0, "step": 1006 }, { "epoch": 0.12810075054064368, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 5.694248199462891, "learning_rate": 4.2645188639253923e-07, "loss": 0.4602, "mean_token_accuracy": 0.8490414619445801, "num_tokens": 38372809.0, "step": 1007 }, { "epoch": 0.12822796081923418, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 5.7468695640563965, "learning_rate": 4.26875794828317e-07, "loss": 0.5188, "mean_token_accuracy": 0.8330071568489075, "num_tokens": 38413312.0, "step": 1008 }, { "epoch": 0.1283551710978247, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.729379177093506, "learning_rate": 4.272997032640949e-07, "loss": 0.5384, "mean_token_accuracy": 0.8279644250869751, "num_tokens": 38457710.0, "step": 1009 }, { "epoch": 0.1284823813764152, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 5.690853118896484, "learning_rate": 4.277236116998728e-07, "loss": 0.4989, "mean_token_accuracy": 0.8402433395385742, "num_tokens": 38499630.0, "step": 1010 }, { "epoch": 0.1286095916550057, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.749075412750244, "learning_rate": 4.281475201356507e-07, "loss": 0.5064, "mean_token_accuracy": 0.8385547399520874, "num_tokens": 38538133.0, "step": 1011 }, { "epoch": 0.12873680193359624, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.746129035949707, "learning_rate": 4.285714285714285e-07, "loss": 0.5382, "mean_token_accuracy": 0.8283088207244873, "num_tokens": 38576816.0, "step": 1012 }, { "epoch": 0.12886401221218674, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 5.727560520172119, "learning_rate": 4.289953370072064e-07, "loss": 0.4781, "mean_token_accuracy": 0.8449018001556396, "num_tokens": 38615270.0, "step": 1013 }, { "epoch": 0.12899122249077727, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.73588228225708, "learning_rate": 4.294192454429843e-07, "loss": 0.5294, "mean_token_accuracy": 0.8292214870452881, "num_tokens": 38654173.0, "step": 1014 }, { "epoch": 0.12911843276936777, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 3.7550926208496094e-06, "grad_norm": 5.6855082511901855, "learning_rate": 4.298431538787622e-07, "loss": 0.4157, "mean_token_accuracy": 0.8669927716255188, "num_tokens": 38695758.0, "step": 1015 }, { "epoch": 0.12924564304795827, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 5.8305768966674805, "learning_rate": 4.3026706231454e-07, "loss": 0.5611, "mean_token_accuracy": 0.8206436038017273, "num_tokens": 38729524.0, "step": 1016 }, { "epoch": 0.1293728533265488, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.7760210037231445, "learning_rate": 4.306909707503179e-07, "loss": 0.5089, "mean_token_accuracy": 0.840303897857666, "num_tokens": 38767500.0, "step": 1017 }, { "epoch": 0.1295000636051393, "ewc_loss": 0.004638671875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.764355659484863, "learning_rate": 4.311148791860958e-07, "loss": 0.4963, "mean_token_accuracy": 0.8392126560211182, "num_tokens": 38798683.0, "step": 1018 }, { "epoch": 0.1296272738837298, "ewc_loss": 0.004638671875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 3.7848949432373047e-06, "grad_norm": 5.758508682250977, "learning_rate": 4.315387876218737e-07, "loss": 0.5452, "mean_token_accuracy": 0.8275349140167236, "num_tokens": 38833653.0, "step": 1019 }, { "epoch": 0.12975448416232033, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.796791076660156, "learning_rate": 4.319626960576515e-07, "loss": 0.5428, "mean_token_accuracy": 0.8238717317581177, "num_tokens": 38869428.0, "step": 1020 }, { "epoch": 0.12988169444091083, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.720977306365967, "learning_rate": 4.323866044934294e-07, "loss": 0.4797, "mean_token_accuracy": 0.8409926295280457, "num_tokens": 38909502.0, "step": 1021 }, { "epoch": 0.13000890471950133, "ewc_loss": 0.004638671875, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 5.718896865844727, "learning_rate": 4.328105129292073e-07, "loss": 0.5035, "mean_token_accuracy": 0.8396272659301758, "num_tokens": 38951730.0, "step": 1022 }, { "epoch": 0.13013611499809186, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.73939847946167, "learning_rate": 4.332344213649852e-07, "loss": 0.4812, "mean_token_accuracy": 0.842044472694397, "num_tokens": 38988981.0, "step": 1023 }, { "epoch": 0.13026332527668236, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.772119045257568, "learning_rate": 4.33658329800763e-07, "loss": 0.4908, "mean_token_accuracy": 0.8429278135299683, "num_tokens": 39026289.0, "step": 1024 }, { "epoch": 0.13039053555527286, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.732489585876465, "learning_rate": 4.340822382365409e-07, "loss": 0.4808, "mean_token_accuracy": 0.8429298400878906, "num_tokens": 39069113.0, "step": 1025 }, { "epoch": 0.13051774583386339, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.736334323883057, "learning_rate": 4.345061466723188e-07, "loss": 0.4651, "mean_token_accuracy": 0.8498891592025757, "num_tokens": 39112991.0, "step": 1026 }, { "epoch": 0.13064495611245389, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.778744220733643, "learning_rate": 4.3493005510809663e-07, "loss": 0.5278, "mean_token_accuracy": 0.832506000995636, "num_tokens": 39149593.0, "step": 1027 }, { "epoch": 0.1307721663910444, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.789658069610596, "learning_rate": 4.353539635438745e-07, "loss": 0.4916, "mean_token_accuracy": 0.8411723971366882, "num_tokens": 39186570.0, "step": 1028 }, { "epoch": 0.13089937666963491, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.859371662139893, "learning_rate": 4.357778719796524e-07, "loss": 0.5549, "mean_token_accuracy": 0.819794774055481, "num_tokens": 39220666.0, "step": 1029 }, { "epoch": 0.13102658694822542, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.720921039581299, "learning_rate": 4.362017804154303e-07, "loss": 0.4734, "mean_token_accuracy": 0.8481924533843994, "num_tokens": 39258372.0, "step": 1030 }, { "epoch": 0.13115379722681592, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.765336990356445, "learning_rate": 4.366256888512081e-07, "loss": 0.469, "mean_token_accuracy": 0.8498009443283081, "num_tokens": 39293647.0, "step": 1031 }, { "epoch": 0.13128100750540644, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.753210067749023, "learning_rate": 4.3704959728698597e-07, "loss": 0.4968, "mean_token_accuracy": 0.838341474533081, "num_tokens": 39332811.0, "step": 1032 }, { "epoch": 0.13140821778399694, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 5.826840877532959, "learning_rate": 4.3747350572276386e-07, "loss": 0.4988, "mean_token_accuracy": 0.8369148969650269, "num_tokens": 39365971.0, "step": 1033 }, { "epoch": 0.13153542806258745, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.790421485900879, "learning_rate": 4.3789741415854176e-07, "loss": 0.5104, "mean_token_accuracy": 0.8365906476974487, "num_tokens": 39403178.0, "step": 1034 }, { "epoch": 0.13166263834117797, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.773758411407471, "learning_rate": 4.383213225943196e-07, "loss": 0.503, "mean_token_accuracy": 0.836323618888855, "num_tokens": 39441303.0, "step": 1035 }, { "epoch": 0.13178984861976847, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.780406951904297, "learning_rate": 4.3874523103009746e-07, "loss": 0.448, "mean_token_accuracy": 0.8554518222808838, "num_tokens": 39477359.0, "step": 1036 }, { "epoch": 0.13191705889835897, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.788763523101807, "learning_rate": 4.3916913946587536e-07, "loss": 0.4579, "mean_token_accuracy": 0.8476597666740417, "num_tokens": 39511687.0, "step": 1037 }, { "epoch": 0.1320442691769495, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.774433612823486, "learning_rate": 4.3959304790165325e-07, "loss": 0.473, "mean_token_accuracy": 0.8447970747947693, "num_tokens": 39547453.0, "step": 1038 }, { "epoch": 0.13217147945554, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.727944374084473, "learning_rate": 4.400169563374311e-07, "loss": 0.4904, "mean_token_accuracy": 0.8408124446868896, "num_tokens": 39588934.0, "step": 1039 }, { "epoch": 0.13229868973413053, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.816929817199707, "learning_rate": 4.4044086477320895e-07, "loss": 0.5812, "mean_token_accuracy": 0.8186309337615967, "num_tokens": 39627638.0, "step": 1040 }, { "epoch": 0.13242590001272103, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 5.821163654327393, "learning_rate": 4.4086477320898685e-07, "loss": 0.4887, "mean_token_accuracy": 0.8401522636413574, "num_tokens": 39661696.0, "step": 1041 }, { "epoch": 0.13255311029131153, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.72619104385376, "learning_rate": 4.4128868164476474e-07, "loss": 0.4981, "mean_token_accuracy": 0.843470573425293, "num_tokens": 39702357.0, "step": 1042 }, { "epoch": 0.13268032056990206, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.785217761993408, "learning_rate": 4.417125900805426e-07, "loss": 0.4821, "mean_token_accuracy": 0.8451467156410217, "num_tokens": 39744960.0, "step": 1043 }, { "epoch": 0.13280753084849256, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.815450191497803, "learning_rate": 4.4213649851632044e-07, "loss": 0.54, "mean_token_accuracy": 0.8305152654647827, "num_tokens": 39785394.0, "step": 1044 }, { "epoch": 0.13293474112708306, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.815251350402832, "learning_rate": 4.4256040695209834e-07, "loss": 0.5214, "mean_token_accuracy": 0.835412859916687, "num_tokens": 39822698.0, "step": 1045 }, { "epoch": 0.1330619514056736, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.85530948638916, "learning_rate": 4.429843153878762e-07, "loss": 0.5177, "mean_token_accuracy": 0.8326214551925659, "num_tokens": 39856874.0, "step": 1046 }, { "epoch": 0.1331891616842641, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.828094005584717, "learning_rate": 4.434082238236541e-07, "loss": 0.5495, "mean_token_accuracy": 0.8237923383712769, "num_tokens": 39899683.0, "step": 1047 }, { "epoch": 0.1333163719628546, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.821719646453857, "learning_rate": 4.4383213225943193e-07, "loss": 0.5282, "mean_token_accuracy": 0.8326901197433472, "num_tokens": 39940565.0, "step": 1048 }, { "epoch": 0.13344358224144512, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.824820518493652, "learning_rate": 4.442560406952098e-07, "loss": 0.528, "mean_token_accuracy": 0.828923225402832, "num_tokens": 39979010.0, "step": 1049 }, { "epoch": 0.13357079252003562, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 5.76363468170166, "learning_rate": 4.4467994913098767e-07, "loss": 0.476, "mean_token_accuracy": 0.8457219004631042, "num_tokens": 40021151.0, "step": 1050 }, { "epoch": 0.13369800279862612, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.830479621887207, "learning_rate": 4.4510385756676557e-07, "loss": 0.5364, "mean_token_accuracy": 0.8269296884536743, "num_tokens": 40059324.0, "step": 1051 }, { "epoch": 0.13382521307721665, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.755762577056885, "learning_rate": 4.455277660025434e-07, "loss": 0.4918, "mean_token_accuracy": 0.8397261500358582, "num_tokens": 40104577.0, "step": 1052 }, { "epoch": 0.13395242335580715, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.883810520172119, "learning_rate": 4.459516744383213e-07, "loss": 0.5154, "mean_token_accuracy": 0.8308308124542236, "num_tokens": 40139734.0, "step": 1053 }, { "epoch": 0.13407963363439765, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.833199501037598, "learning_rate": 4.4637558287409916e-07, "loss": 0.5135, "mean_token_accuracy": 0.8344751596450806, "num_tokens": 40180724.0, "step": 1054 }, { "epoch": 0.13420684391298818, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.790544033050537, "learning_rate": 4.4679949130987706e-07, "loss": 0.5126, "mean_token_accuracy": 0.840364933013916, "num_tokens": 40220475.0, "step": 1055 }, { "epoch": 0.13433405419157868, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.831809997558594, "learning_rate": 4.472233997456549e-07, "loss": 0.5217, "mean_token_accuracy": 0.8325462341308594, "num_tokens": 40258208.0, "step": 1056 }, { "epoch": 0.13446126447016918, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.821534156799316, "learning_rate": 4.476473081814328e-07, "loss": 0.5071, "mean_token_accuracy": 0.8351930379867554, "num_tokens": 40299127.0, "step": 1057 }, { "epoch": 0.1345884747487597, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.807558059692383, "learning_rate": 4.4807121661721065e-07, "loss": 0.466, "mean_token_accuracy": 0.847161054611206, "num_tokens": 40339331.0, "step": 1058 }, { "epoch": 0.1347156850273502, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.818914413452148, "learning_rate": 4.4849512505298855e-07, "loss": 0.5095, "mean_token_accuracy": 0.8355880975723267, "num_tokens": 40379386.0, "step": 1059 }, { "epoch": 0.1348428953059407, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 5.811965465545654, "learning_rate": 4.489190334887664e-07, "loss": 0.5462, "mean_token_accuracy": 0.8314412832260132, "num_tokens": 40419868.0, "step": 1060 }, { "epoch": 0.13497010558453124, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.890603065490723, "learning_rate": 4.493429419245443e-07, "loss": 0.5149, "mean_token_accuracy": 0.8364571332931519, "num_tokens": 40455346.0, "step": 1061 }, { "epoch": 0.13509731586312174, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.888159275054932, "learning_rate": 4.4976685036032214e-07, "loss": 0.4735, "mean_token_accuracy": 0.8455995917320251, "num_tokens": 40490087.0, "step": 1062 }, { "epoch": 0.13522452614171224, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.874096870422363, "learning_rate": 4.5019075879610004e-07, "loss": 0.5313, "mean_token_accuracy": 0.8268123865127563, "num_tokens": 40524799.0, "step": 1063 }, { "epoch": 0.13535173642030277, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.851675510406494, "learning_rate": 4.506146672318779e-07, "loss": 0.5039, "mean_token_accuracy": 0.8371754884719849, "num_tokens": 40563740.0, "step": 1064 }, { "epoch": 0.13547894669889327, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.864901065826416, "learning_rate": 4.5103857566765573e-07, "loss": 0.5579, "mean_token_accuracy": 0.8264778852462769, "num_tokens": 40604743.0, "step": 1065 }, { "epoch": 0.1356061569774838, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.885894298553467, "learning_rate": 4.5146248410343363e-07, "loss": 0.4631, "mean_token_accuracy": 0.8503403663635254, "num_tokens": 40640909.0, "step": 1066 }, { "epoch": 0.1357333672560743, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.82905912399292, "learning_rate": 4.5188639253921153e-07, "loss": 0.4716, "mean_token_accuracy": 0.8477888107299805, "num_tokens": 40678930.0, "step": 1067 }, { "epoch": 0.1358605775346648, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.894052505493164, "learning_rate": 4.523103009749894e-07, "loss": 0.4659, "mean_token_accuracy": 0.8470329642295837, "num_tokens": 40715609.0, "step": 1068 }, { "epoch": 0.13598778781325532, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.819043159484863, "learning_rate": 4.527342094107672e-07, "loss": 0.4907, "mean_token_accuracy": 0.8407394886016846, "num_tokens": 40760533.0, "step": 1069 }, { "epoch": 0.13611499809184582, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.866238117218018, "learning_rate": 4.531581178465451e-07, "loss": 0.5179, "mean_token_accuracy": 0.8341139554977417, "num_tokens": 40800039.0, "step": 1070 }, { "epoch": 0.13624220837043632, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.887661933898926, "learning_rate": 4.53582026282323e-07, "loss": 0.4806, "mean_token_accuracy": 0.8449245691299438, "num_tokens": 40840207.0, "step": 1071 }, { "epoch": 0.13636941864902685, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.925565242767334, "learning_rate": 4.5400593471810087e-07, "loss": 0.5624, "mean_token_accuracy": 0.8195964097976685, "num_tokens": 40879559.0, "step": 1072 }, { "epoch": 0.13649662892761735, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.8898186683654785, "learning_rate": 4.544298431538787e-07, "loss": 0.4899, "mean_token_accuracy": 0.8417088985443115, "num_tokens": 40919608.0, "step": 1073 }, { "epoch": 0.13662383920620785, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.920863151550293, "learning_rate": 4.548537515896566e-07, "loss": 0.5504, "mean_token_accuracy": 0.8234808444976807, "num_tokens": 40953648.0, "step": 1074 }, { "epoch": 0.13675104948479838, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.941351413726807, "learning_rate": 4.552776600254345e-07, "loss": 0.5224, "mean_token_accuracy": 0.8330569267272949, "num_tokens": 40989675.0, "step": 1075 }, { "epoch": 0.13687825976338888, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.8856282234191895, "learning_rate": 4.5570156846121236e-07, "loss": 0.4738, "mean_token_accuracy": 0.8470573425292969, "num_tokens": 41027992.0, "step": 1076 }, { "epoch": 0.13700547004197938, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.891098499298096, "learning_rate": 4.561254768969902e-07, "loss": 0.4729, "mean_token_accuracy": 0.8441423177719116, "num_tokens": 41066925.0, "step": 1077 }, { "epoch": 0.1371326803205699, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.905167102813721, "learning_rate": 4.565493853327681e-07, "loss": 0.482, "mean_token_accuracy": 0.8444938659667969, "num_tokens": 41105090.0, "step": 1078 }, { "epoch": 0.1372598905991604, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.93950080871582, "learning_rate": 4.56973293768546e-07, "loss": 0.5245, "mean_token_accuracy": 0.834796667098999, "num_tokens": 41141924.0, "step": 1079 }, { "epoch": 0.1373871008777509, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.899685859680176, "learning_rate": 4.573972022043238e-07, "loss": 0.5855, "mean_token_accuracy": 0.8133052587509155, "num_tokens": 41181512.0, "step": 1080 }, { "epoch": 0.13751431115634144, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.856581211090088, "learning_rate": 4.578211106401017e-07, "loss": 0.5336, "mean_token_accuracy": 0.8283716440200806, "num_tokens": 41224015.0, "step": 1081 }, { "epoch": 0.13764152143493194, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 5.889354705810547, "learning_rate": 4.582450190758796e-07, "loss": 0.5354, "mean_token_accuracy": 0.8305872082710266, "num_tokens": 41265568.0, "step": 1082 }, { "epoch": 0.13776873171352244, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.875610828399658, "learning_rate": 4.586689275116575e-07, "loss": 0.4493, "mean_token_accuracy": 0.8544760942459106, "num_tokens": 41303595.0, "step": 1083 }, { "epoch": 0.13789594199211297, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.896464824676514, "learning_rate": 4.590928359474353e-07, "loss": 0.5526, "mean_token_accuracy": 0.8251062035560608, "num_tokens": 41340119.0, "step": 1084 }, { "epoch": 0.13802315227070347, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.900921821594238, "learning_rate": 4.595167443832132e-07, "loss": 0.5042, "mean_token_accuracy": 0.840883731842041, "num_tokens": 41380184.0, "step": 1085 }, { "epoch": 0.13815036254929397, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.905335903167725, "learning_rate": 4.599406528189911e-07, "loss": 0.4473, "mean_token_accuracy": 0.855106770992279, "num_tokens": 41419038.0, "step": 1086 }, { "epoch": 0.1382775728278845, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.907649993896484, "learning_rate": 4.60364561254769e-07, "loss": 0.4904, "mean_token_accuracy": 0.838543713092804, "num_tokens": 41455767.0, "step": 1087 }, { "epoch": 0.138404783106475, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.8905720710754395, "learning_rate": 4.607884696905468e-07, "loss": 0.4873, "mean_token_accuracy": 0.8413307666778564, "num_tokens": 41497493.0, "step": 1088 }, { "epoch": 0.1385319933850655, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.916534423828125, "learning_rate": 4.612123781263247e-07, "loss": 0.5189, "mean_token_accuracy": 0.8331390023231506, "num_tokens": 41542940.0, "step": 1089 }, { "epoch": 0.13865920366365603, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.906485557556152, "learning_rate": 4.616362865621026e-07, "loss": 0.4195, "mean_token_accuracy": 0.8604656457901001, "num_tokens": 41579302.0, "step": 1090 }, { "epoch": 0.13878641394224653, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.936972618103027, "learning_rate": 4.620601949978805e-07, "loss": 0.5177, "mean_token_accuracy": 0.8351479768753052, "num_tokens": 41617555.0, "step": 1091 }, { "epoch": 0.13891362422083706, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.010808944702148, "learning_rate": 4.6248410343365827e-07, "loss": 0.5696, "mean_token_accuracy": 0.8179984092712402, "num_tokens": 41654660.0, "step": 1092 }, { "epoch": 0.13904083449942756, "ewc_loss": 0.0048828125, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.9134016036987305, "learning_rate": 4.6290801186943617e-07, "loss": 0.4764, "mean_token_accuracy": 0.8453797101974487, "num_tokens": 41693773.0, "step": 1093 }, { "epoch": 0.13916804477801806, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.954039573669434, "learning_rate": 4.6333192030521407e-07, "loss": 0.4908, "mean_token_accuracy": 0.8441152572631836, "num_tokens": 41728864.0, "step": 1094 }, { "epoch": 0.13929525505660859, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.898492336273193, "learning_rate": 4.6375582874099196e-07, "loss": 0.465, "mean_token_accuracy": 0.8498154878616333, "num_tokens": 41770596.0, "step": 1095 }, { "epoch": 0.1394224653351991, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.94131326675415, "learning_rate": 4.6417973717676976e-07, "loss": 0.4828, "mean_token_accuracy": 0.8465547561645508, "num_tokens": 41806585.0, "step": 1096 }, { "epoch": 0.1395496756137896, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 5.947323322296143, "learning_rate": 4.6460364561254766e-07, "loss": 0.5116, "mean_token_accuracy": 0.8341696262359619, "num_tokens": 41845488.0, "step": 1097 }, { "epoch": 0.13967688589238011, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.9839582443237305, "learning_rate": 4.6502755404832556e-07, "loss": 0.478, "mean_token_accuracy": 0.844208300113678, "num_tokens": 41880674.0, "step": 1098 }, { "epoch": 0.13980409617097062, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.985847473144531, "learning_rate": 4.654514624841034e-07, "loss": 0.4924, "mean_token_accuracy": 0.840703010559082, "num_tokens": 41913020.0, "step": 1099 }, { "epoch": 0.13993130644956112, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.957752227783203, "learning_rate": 4.6587537091988125e-07, "loss": 0.4979, "mean_token_accuracy": 0.8407496213912964, "num_tokens": 41948483.0, "step": 1100 }, { "epoch": 0.14005851672815164, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 6.002047538757324, "learning_rate": 4.6629927935565915e-07, "loss": 0.4699, "mean_token_accuracy": 0.8473066091537476, "num_tokens": 41985659.0, "step": 1101 }, { "epoch": 0.14018572700674214, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 5.948823928833008, "learning_rate": 4.6672318779143705e-07, "loss": 0.5655, "mean_token_accuracy": 0.819491982460022, "num_tokens": 42027746.0, "step": 1102 }, { "epoch": 0.14031293728533265, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 6.0010576248168945, "learning_rate": 4.671470962272149e-07, "loss": 0.5211, "mean_token_accuracy": 0.8317404985427856, "num_tokens": 42060903.0, "step": 1103 }, { "epoch": 0.14044014756392317, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 5.941601276397705, "learning_rate": 4.6757100466299274e-07, "loss": 0.5175, "mean_token_accuracy": 0.8371507525444031, "num_tokens": 42102722.0, "step": 1104 }, { "epoch": 0.14056735784251367, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 9.5367431640625e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 5.933562278747559, "learning_rate": 4.6799491309877064e-07, "loss": 0.526, "mean_token_accuracy": 0.8288871049880981, "num_tokens": 42143418.0, "step": 1105 }, { "epoch": 0.14069456812110417, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.5367431640625e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 5.973007678985596, "learning_rate": 4.6841882153454854e-07, "loss": 0.4655, "mean_token_accuracy": 0.8451521396636963, "num_tokens": 42177313.0, "step": 1106 }, { "epoch": 0.1408217783996947, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 9.5367431640625e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 5.973716735839844, "learning_rate": 4.688427299703264e-07, "loss": 0.446, "mean_token_accuracy": 0.8522483110427856, "num_tokens": 42214158.0, "step": 1107 }, { "epoch": 0.1409489886782852, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 9.5367431640625e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 5.9427361488342285, "learning_rate": 4.6926663840610423e-07, "loss": 0.504, "mean_token_accuracy": 0.8391066193580627, "num_tokens": 42254189.0, "step": 1108 }, { "epoch": 0.1410761989568757, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 9.5367431640625e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 5.965338706970215, "learning_rate": 4.6969054684188213e-07, "loss": 0.5257, "mean_token_accuracy": 0.8283756971359253, "num_tokens": 42296853.0, "step": 1109 }, { "epoch": 0.14120340923546623, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.026130199432373, "learning_rate": 4.7011445527766003e-07, "loss": 0.4971, "mean_token_accuracy": 0.8370525240898132, "num_tokens": 42331983.0, "step": 1110 }, { "epoch": 0.14133061951405673, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.036466598510742, "learning_rate": 4.7053836371343787e-07, "loss": 0.506, "mean_token_accuracy": 0.8387897610664368, "num_tokens": 42370496.0, "step": 1111 }, { "epoch": 0.14145782979264723, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 5.966765403747559, "learning_rate": 4.709622721492157e-07, "loss": 0.4839, "mean_token_accuracy": 0.844089925289154, "num_tokens": 42406489.0, "step": 1112 }, { "epoch": 0.14158504007123776, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 6.03069543838501, "learning_rate": 4.713861805849936e-07, "loss": 0.4844, "mean_token_accuracy": 0.8419486284255981, "num_tokens": 42440679.0, "step": 1113 }, { "epoch": 0.14171225034982826, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.029045104980469, "learning_rate": 4.718100890207715e-07, "loss": 0.5567, "mean_token_accuracy": 0.8215541839599609, "num_tokens": 42476699.0, "step": 1114 }, { "epoch": 0.1418394606284188, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.019526958465576, "learning_rate": 4.7223399745654936e-07, "loss": 0.5127, "mean_token_accuracy": 0.8345460295677185, "num_tokens": 42513348.0, "step": 1115 }, { "epoch": 0.1419666709070093, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 5.986627101898193, "learning_rate": 4.726579058923272e-07, "loss": 0.5048, "mean_token_accuracy": 0.8396267294883728, "num_tokens": 42550633.0, "step": 1116 }, { "epoch": 0.1420938811855998, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.003553867340088, "learning_rate": 4.730818143281051e-07, "loss": 0.5292, "mean_token_accuracy": 0.8321356177330017, "num_tokens": 42593649.0, "step": 1117 }, { "epoch": 0.14222109146419032, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.013800144195557, "learning_rate": 4.7350572276388295e-07, "loss": 0.4761, "mean_token_accuracy": 0.8460149168968201, "num_tokens": 42632379.0, "step": 1118 }, { "epoch": 0.14234830174278082, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.016264915466309, "learning_rate": 4.7392963119966085e-07, "loss": 0.5027, "mean_token_accuracy": 0.8358671069145203, "num_tokens": 42670319.0, "step": 1119 }, { "epoch": 0.14247551202137132, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 5.990714073181152, "learning_rate": 4.7435353963543875e-07, "loss": 0.4959, "mean_token_accuracy": 0.8421945571899414, "num_tokens": 42712680.0, "step": 1120 }, { "epoch": 0.14260272229996185, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.0313801765441895, "learning_rate": 4.747774480712166e-07, "loss": 0.513, "mean_token_accuracy": 0.8374277949333191, "num_tokens": 42748241.0, "step": 1121 }, { "epoch": 0.14272993257855235, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.039247989654541, "learning_rate": 4.7520135650699444e-07, "loss": 0.5328, "mean_token_accuracy": 0.8302850127220154, "num_tokens": 42783208.0, "step": 1122 }, { "epoch": 0.14285714285714285, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.057292461395264, "learning_rate": 4.7562526494277234e-07, "loss": 0.5079, "mean_token_accuracy": 0.8388113975524902, "num_tokens": 42818570.0, "step": 1123 }, { "epoch": 0.14298435313573338, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.013481140136719, "learning_rate": 4.7604917337855024e-07, "loss": 0.4467, "mean_token_accuracy": 0.8522971272468567, "num_tokens": 42852891.0, "step": 1124 }, { "epoch": 0.14311156341432388, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.051042556762695, "learning_rate": 4.764730818143281e-07, "loss": 0.4965, "mean_token_accuracy": 0.8336466550827026, "num_tokens": 42887359.0, "step": 1125 }, { "epoch": 0.14323877369291438, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.013698577880859, "learning_rate": 4.768969902501059e-07, "loss": 0.5493, "mean_token_accuracy": 0.8236426711082458, "num_tokens": 42929308.0, "step": 1126 }, { "epoch": 0.1433659839715049, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 5.992674350738525, "learning_rate": 4.773208986858838e-07, "loss": 0.499, "mean_token_accuracy": 0.8361718654632568, "num_tokens": 42969373.0, "step": 1127 }, { "epoch": 0.1434931942500954, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 5.991809844970703, "learning_rate": 4.777448071216617e-07, "loss": 0.504, "mean_token_accuracy": 0.8385668396949768, "num_tokens": 43010133.0, "step": 1128 }, { "epoch": 0.1436204045286859, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.0246734619140625, "learning_rate": 4.781687155574396e-07, "loss": 0.5289, "mean_token_accuracy": 0.8296384811401367, "num_tokens": 43053743.0, "step": 1129 }, { "epoch": 0.14374761480727644, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.0114946365356445, "learning_rate": 4.785926239932175e-07, "loss": 0.4695, "mean_token_accuracy": 0.8470019698143005, "num_tokens": 43094125.0, "step": 1130 }, { "epoch": 0.14387482508586694, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.027503967285156, "learning_rate": 4.790165324289953e-07, "loss": 0.4965, "mean_token_accuracy": 0.8373904824256897, "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": 6.088526248931885, "learning_rate": 4.794404408647732e-07, "loss": 0.5103, "mean_token_accuracy": 0.8376268148422241, "num_tokens": 43169036.0, "step": 1132 }, { "epoch": 0.14412924564304797, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.020599365234375, "learning_rate": 4.798643493005511e-07, "loss": 0.4412, "mean_token_accuracy": 0.8547101616859436, "num_tokens": 43209397.0, "step": 1133 }, { "epoch": 0.14425645592163847, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.025384902954102, "learning_rate": 4.80288257736329e-07, "loss": 0.5384, "mean_token_accuracy": 0.8292557001113892, "num_tokens": 43246788.0, "step": 1134 }, { "epoch": 0.14438366620022897, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.0233259201049805, "learning_rate": 4.807121661721068e-07, "loss": 0.451, "mean_token_accuracy": 0.85343337059021, "num_tokens": 43282774.0, "step": 1135 }, { "epoch": 0.1445108764788195, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.048872470855713, "learning_rate": 4.811360746078847e-07, "loss": 0.5066, "mean_token_accuracy": 0.8407851457595825, "num_tokens": 43322303.0, "step": 1136 }, { "epoch": 0.14463808675741, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 6.105279922485352, "learning_rate": 4.815599830436625e-07, "loss": 0.5004, "mean_token_accuracy": 0.838245153427124, "num_tokens": 43354611.0, "step": 1137 }, { "epoch": 0.1447652970360005, "ewc_loss": 0.005126953125, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.0662665367126465, "learning_rate": 4.819838914794405e-07, "loss": 0.4821, "mean_token_accuracy": 0.8447365760803223, "num_tokens": 43390640.0, "step": 1138 }, { "epoch": 0.14489250731459102, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 6.020385265350342, "learning_rate": 4.824077999152183e-07, "loss": 0.5019, "mean_token_accuracy": 0.8338304758071899, "num_tokens": 43424513.0, "step": 1139 }, { "epoch": 0.14501971759318152, "ewc_loss": 0.005126953125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 6.055619716644287, "learning_rate": 4.828317083509962e-07, "loss": 0.5078, "mean_token_accuracy": 0.8354668021202087, "num_tokens": 43461699.0, "step": 1140 }, { "epoch": 0.14514692787177205, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.173074722290039, "learning_rate": 4.83255616786774e-07, "loss": 0.4819, "mean_token_accuracy": 0.8481189012527466, "num_tokens": 43494681.0, "step": 1141 }, { "epoch": 0.14527413815036255, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.045533657073975, "learning_rate": 4.83679525222552e-07, "loss": 0.4578, "mean_token_accuracy": 0.8509320020675659, "num_tokens": 43531318.0, "step": 1142 }, { "epoch": 0.14540134842895305, "ewc_loss": 0.005126953125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 6.023292541503906, "learning_rate": 4.841034336583298e-07, "loss": 0.4742, "mean_token_accuracy": 0.8510741591453552, "num_tokens": 43569991.0, "step": 1143 }, { "epoch": 0.14552855870754358, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.05891227722168, "learning_rate": 4.845273420941076e-07, "loss": 0.4619, "mean_token_accuracy": 0.8517163395881653, "num_tokens": 43609978.0, "step": 1144 }, { "epoch": 0.14565576898613408, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.1172943115234375, "learning_rate": 4.849512505298855e-07, "loss": 0.5103, "mean_token_accuracy": 0.8303350806236267, "num_tokens": 43640419.0, "step": 1145 }, { "epoch": 0.14578297926472458, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.024559020996094, "learning_rate": 4.853751589656634e-07, "loss": 0.4717, "mean_token_accuracy": 0.8479756116867065, "num_tokens": 43682408.0, "step": 1146 }, { "epoch": 0.1459101895433151, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.033487796783447, "learning_rate": 4.857990674014413e-07, "loss": 0.5098, "mean_token_accuracy": 0.8348962664604187, "num_tokens": 43726158.0, "step": 1147 }, { "epoch": 0.1460373998219056, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.072062969207764, "learning_rate": 4.862229758372191e-07, "loss": 0.539, "mean_token_accuracy": 0.824869692325592, "num_tokens": 43766466.0, "step": 1148 }, { "epoch": 0.1461646101004961, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.075615406036377, "learning_rate": 4.86646884272997e-07, "loss": 0.5265, "mean_token_accuracy": 0.8278582096099854, "num_tokens": 43807255.0, "step": 1149 }, { "epoch": 0.14629182037908664, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.115794658660889, "learning_rate": 4.870707927087749e-07, "loss": 0.5053, "mean_token_accuracy": 0.8366640210151672, "num_tokens": 43845419.0, "step": 1150 }, { "epoch": 0.14641903065767714, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.116551876068115, "learning_rate": 4.874947011445528e-07, "loss": 0.5136, "mean_token_accuracy": 0.8366727828979492, "num_tokens": 43882942.0, "step": 1151 }, { "epoch": 0.14654624093626764, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.023918628692627, "learning_rate": 4.879186095803306e-07, "loss": 0.4343, "mean_token_accuracy": 0.858698308467865, "num_tokens": 43923109.0, "step": 1152 }, { "epoch": 0.14667345121485817, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.112142562866211, "learning_rate": 4.883425180161085e-07, "loss": 0.4712, "mean_token_accuracy": 0.846340000629425, "num_tokens": 43959752.0, "step": 1153 }, { "epoch": 0.14680066149344867, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.142483234405518, "learning_rate": 4.887664264518864e-07, "loss": 0.5184, "mean_token_accuracy": 0.8379865884780884, "num_tokens": 43996247.0, "step": 1154 }, { "epoch": 0.14692787177203917, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.067031383514404, "learning_rate": 4.891903348876643e-07, "loss": 0.5016, "mean_token_accuracy": 0.8419232368469238, "num_tokens": 44035751.0, "step": 1155 }, { "epoch": 0.1470550820506297, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.062411785125732, "learning_rate": 4.896142433234421e-07, "loss": 0.536, "mean_token_accuracy": 0.8341379165649414, "num_tokens": 44073225.0, "step": 1156 }, { "epoch": 0.1471822923292202, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.134648323059082, "learning_rate": 4.9003815175922e-07, "loss": 0.5389, "mean_token_accuracy": 0.8325982689857483, "num_tokens": 44111761.0, "step": 1157 }, { "epoch": 0.1473095026078107, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.056484222412109, "learning_rate": 4.904620601949979e-07, "loss": 0.4899, "mean_token_accuracy": 0.8417183756828308, "num_tokens": 44151311.0, "step": 1158 }, { "epoch": 0.14743671288640123, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 6.1318864822387695, "learning_rate": 4.908859686307758e-07, "loss": 0.5343, "mean_token_accuracy": 0.8311312198638916, "num_tokens": 44184623.0, "step": 1159 }, { "epoch": 0.14756392316499173, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.101343154907227, "learning_rate": 4.913098770665536e-07, "loss": 0.5206, "mean_token_accuracy": 0.8364394903182983, "num_tokens": 44221753.0, "step": 1160 }, { "epoch": 0.14769113344358223, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.0826921463012695, "learning_rate": 4.917337855023314e-07, "loss": 0.4737, "mean_token_accuracy": 0.8430840969085693, "num_tokens": 44262542.0, "step": 1161 }, { "epoch": 0.14781834372217276, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.123533725738525, "learning_rate": 4.921576939381094e-07, "loss": 0.4963, "mean_token_accuracy": 0.8448676466941833, "num_tokens": 44300294.0, "step": 1162 }, { "epoch": 0.14794555400076326, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.128901481628418, "learning_rate": 4.925816023738872e-07, "loss": 0.5417, "mean_token_accuracy": 0.8260762095451355, "num_tokens": 44338476.0, "step": 1163 }, { "epoch": 0.14807276427935376, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.1793389320373535, "learning_rate": 4.930055108096651e-07, "loss": 0.4436, "mean_token_accuracy": 0.8555219173431396, "num_tokens": 44367437.0, "step": 1164 }, { "epoch": 0.1481999745579443, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.151729106903076, "learning_rate": 4.934294192454429e-07, "loss": 0.5546, "mean_token_accuracy": 0.823028564453125, "num_tokens": 44402636.0, "step": 1165 }, { "epoch": 0.1483271848365348, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.1419596672058105, "learning_rate": 4.938533276812209e-07, "loss": 0.4967, "mean_token_accuracy": 0.8410837650299072, "num_tokens": 44438187.0, "step": 1166 }, { "epoch": 0.14845439511512531, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.087555885314941, "learning_rate": 4.942772361169987e-07, "loss": 0.4198, "mean_token_accuracy": 0.8587483167648315, "num_tokens": 44478345.0, "step": 1167 }, { "epoch": 0.14858160539371582, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.092111110687256, "learning_rate": 4.947011445527766e-07, "loss": 0.4994, "mean_token_accuracy": 0.8396961092948914, "num_tokens": 44520272.0, "step": 1168 }, { "epoch": 0.14870881567230632, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 6.117664337158203, "learning_rate": 4.951250529885544e-07, "loss": 0.4513, "mean_token_accuracy": 0.855726957321167, "num_tokens": 44561058.0, "step": 1169 }, { "epoch": 0.14883602595089684, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.170928955078125, "learning_rate": 4.955489614243324e-07, "loss": 0.5837, "mean_token_accuracy": 0.8115016222000122, "num_tokens": 44599371.0, "step": 1170 }, { "epoch": 0.14896323622948734, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.136025905609131, "learning_rate": 4.959728698601102e-07, "loss": 0.4947, "mean_token_accuracy": 0.83228999376297, "num_tokens": 44640499.0, "step": 1171 }, { "epoch": 0.14909044650807785, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 6.08797025680542, "learning_rate": 4.963967782958881e-07, "loss": 0.4816, "mean_token_accuracy": 0.8448299765586853, "num_tokens": 44683164.0, "step": 1172 }, { "epoch": 0.14921765678666837, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.127914905548096, "learning_rate": 4.968206867316659e-07, "loss": 0.4935, "mean_token_accuracy": 0.8394964933395386, "num_tokens": 44721938.0, "step": 1173 }, { "epoch": 0.14934486706525887, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.211418628692627, "learning_rate": 4.972445951674439e-07, "loss": 0.4648, "mean_token_accuracy": 0.8473013639450073, "num_tokens": 44753506.0, "step": 1174 }, { "epoch": 0.14947207734384937, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.1758246421813965, "learning_rate": 4.976685036032216e-07, "loss": 0.5558, "mean_token_accuracy": 0.82196044921875, "num_tokens": 44792170.0, "step": 1175 }, { "epoch": 0.1495992876224399, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 6.0982584953308105, "learning_rate": 4.980924120389996e-07, "loss": 0.4327, "mean_token_accuracy": 0.859835147857666, "num_tokens": 44830733.0, "step": 1176 }, { "epoch": 0.1497264979010304, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 6.126740455627441, "learning_rate": 4.985163204747774e-07, "loss": 0.4462, "mean_token_accuracy": 0.855603814125061, "num_tokens": 44869743.0, "step": 1177 }, { "epoch": 0.1498537081796209, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.184301853179932, "learning_rate": 4.989402289105554e-07, "loss": 0.493, "mean_token_accuracy": 0.8461288213729858, "num_tokens": 44913358.0, "step": 1178 }, { "epoch": 0.14998091845821143, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.207702159881592, "learning_rate": 4.993641373463331e-07, "loss": 0.552, "mean_token_accuracy": 0.8230180144309998, "num_tokens": 44949963.0, "step": 1179 }, { "epoch": 0.15010812873680193, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.160505294799805, "learning_rate": 4.997880457821111e-07, "loss": 0.507, "mean_token_accuracy": 0.8351839184761047, "num_tokens": 44988569.0, "step": 1180 }, { "epoch": 0.15023533901539243, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.114008903503418, "learning_rate": 5.002119542178889e-07, "loss": 0.484, "mean_token_accuracy": 0.8425471186637878, "num_tokens": 45032634.0, "step": 1181 }, { "epoch": 0.15036254929398296, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.1779985427856445, "learning_rate": 5.006358626536667e-07, "loss": 0.4527, "mean_token_accuracy": 0.8540617227554321, "num_tokens": 45069685.0, "step": 1182 }, { "epoch": 0.15048975957257346, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.165994644165039, "learning_rate": 5.010597710894446e-07, "loss": 0.4755, "mean_token_accuracy": 0.8451996445655823, "num_tokens": 45107481.0, "step": 1183 }, { "epoch": 0.15061696985116396, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.162552833557129, "learning_rate": 5.014836795252225e-07, "loss": 0.4597, "mean_token_accuracy": 0.8499320149421692, "num_tokens": 45145747.0, "step": 1184 }, { "epoch": 0.1507441801297545, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.154494285583496, "learning_rate": 5.019075879610004e-07, "loss": 0.4671, "mean_token_accuracy": 0.85148686170578, "num_tokens": 45183140.0, "step": 1185 }, { "epoch": 0.150871390408345, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.154510974884033, "learning_rate": 5.023314963967783e-07, "loss": 0.5005, "mean_token_accuracy": 0.8379645347595215, "num_tokens": 45224911.0, "step": 1186 }, { "epoch": 0.1509986006869355, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.224445819854736, "learning_rate": 5.027554048325562e-07, "loss": 0.5553, "mean_token_accuracy": 0.8222665786743164, "num_tokens": 45257388.0, "step": 1187 }, { "epoch": 0.15112581096552602, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.18388557434082, "learning_rate": 5.03179313268334e-07, "loss": 0.5415, "mean_token_accuracy": 0.8270012140274048, "num_tokens": 45297876.0, "step": 1188 }, { "epoch": 0.15125302124411652, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.191375255584717, "learning_rate": 5.036032217041119e-07, "loss": 0.5164, "mean_token_accuracy": 0.831902027130127, "num_tokens": 45334040.0, "step": 1189 }, { "epoch": 0.15138023152270705, "ewc_loss": 0.00537109375, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 6.159334659576416, "learning_rate": 5.040271301398897e-07, "loss": 0.5, "mean_token_accuracy": 0.835696816444397, "num_tokens": 45375985.0, "step": 1190 }, { "epoch": 0.15150744180129755, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.138723850250244, "learning_rate": 5.044510385756676e-07, "loss": 0.4688, "mean_token_accuracy": 0.8452937006950378, "num_tokens": 45419951.0, "step": 1191 }, { "epoch": 0.15163465207988805, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.178598403930664, "learning_rate": 5.048749470114455e-07, "loss": 0.4815, "mean_token_accuracy": 0.8451799154281616, "num_tokens": 45456988.0, "step": 1192 }, { "epoch": 0.15176186235847858, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.278730869293213, "learning_rate": 5.052988554472234e-07, "loss": 0.5001, "mean_token_accuracy": 0.8371899724006653, "num_tokens": 45488226.0, "step": 1193 }, { "epoch": 0.15188907263706908, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.178174018859863, "learning_rate": 5.057227638830013e-07, "loss": 0.4987, "mean_token_accuracy": 0.8397109508514404, "num_tokens": 45524661.0, "step": 1194 }, { "epoch": 0.15201628291565958, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 6.219667434692383, "learning_rate": 5.061466723187792e-07, "loss": 0.501, "mean_token_accuracy": 0.8401470184326172, "num_tokens": 45563491.0, "step": 1195 }, { "epoch": 0.1521434931942501, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.233740329742432, "learning_rate": 5.065705807545569e-07, "loss": 0.4704, "mean_token_accuracy": 0.8495969176292419, "num_tokens": 45596771.0, "step": 1196 }, { "epoch": 0.1522707034728406, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.190141201019287, "learning_rate": 5.069944891903349e-07, "loss": 0.4876, "mean_token_accuracy": 0.8413569927215576, "num_tokens": 45640049.0, "step": 1197 }, { "epoch": 0.1523979137514311, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.19430685043335, "learning_rate": 5.074183976261127e-07, "loss": 0.4894, "mean_token_accuracy": 0.8406583666801453, "num_tokens": 45677375.0, "step": 1198 }, { "epoch": 0.15252512403002164, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.249449729919434, "learning_rate": 5.078423060618906e-07, "loss": 0.4933, "mean_token_accuracy": 0.8426191806793213, "num_tokens": 45719253.0, "step": 1199 }, { "epoch": 0.15265233430861214, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.205742359161377, "learning_rate": 5.082662144976685e-07, "loss": 0.4889, "mean_token_accuracy": 0.8382990956306458, "num_tokens": 45756065.0, "step": 1200 }, { "epoch": 0.15277954458720264, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.203292369842529, "learning_rate": 5.086901229334464e-07, "loss": 0.4784, "mean_token_accuracy": 0.8472594022750854, "num_tokens": 45791707.0, "step": 1201 }, { "epoch": 0.15290675486579317, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.220393657684326, "learning_rate": 5.091140313692243e-07, "loss": 0.4928, "mean_token_accuracy": 0.8393131494522095, "num_tokens": 45828464.0, "step": 1202 }, { "epoch": 0.15303396514438367, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.191131114959717, "learning_rate": 5.095379398050022e-07, "loss": 0.4756, "mean_token_accuracy": 0.8450838923454285, "num_tokens": 45867046.0, "step": 1203 }, { "epoch": 0.15316117542297417, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.2821431159973145, "learning_rate": 5.099618482407799e-07, "loss": 0.5063, "mean_token_accuracy": 0.8372638821601868, "num_tokens": 45900530.0, "step": 1204 }, { "epoch": 0.1532883857015647, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.2066826820373535, "learning_rate": 5.103857566765578e-07, "loss": 0.4838, "mean_token_accuracy": 0.8445221185684204, "num_tokens": 45940002.0, "step": 1205 }, { "epoch": 0.1534155959801552, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 6.241446018218994, "learning_rate": 5.108096651123357e-07, "loss": 0.4466, "mean_token_accuracy": 0.8541410565376282, "num_tokens": 45975299.0, "step": 1206 }, { "epoch": 0.1535428062587457, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.259933948516846, "learning_rate": 5.112335735481135e-07, "loss": 0.4858, "mean_token_accuracy": 0.8412758111953735, "num_tokens": 46012693.0, "step": 1207 }, { "epoch": 0.15367001653733622, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.182723045349121, "learning_rate": 5.116574819838915e-07, "loss": 0.3917, "mean_token_accuracy": 0.8716939687728882, "num_tokens": 46051558.0, "step": 1208 }, { "epoch": 0.15379722681592672, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 6.2482686042785645, "learning_rate": 5.120813904196693e-07, "loss": 0.4749, "mean_token_accuracy": 0.8471970558166504, "num_tokens": 46087173.0, "step": 1209 }, { "epoch": 0.15392443709451722, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.289381980895996, "learning_rate": 5.125052988554473e-07, "loss": 0.5304, "mean_token_accuracy": 0.8325574398040771, "num_tokens": 46120482.0, "step": 1210 }, { "epoch": 0.15405164737310775, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.2466020584106445, "learning_rate": 5.12929207291225e-07, "loss": 0.4773, "mean_token_accuracy": 0.8439375162124634, "num_tokens": 46154478.0, "step": 1211 }, { "epoch": 0.15417885765169825, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 6.196451663970947, "learning_rate": 5.133531157270029e-07, "loss": 0.4546, "mean_token_accuracy": 0.8522951006889343, "num_tokens": 46195012.0, "step": 1212 }, { "epoch": 0.15430606793028875, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.249269008636475, "learning_rate": 5.137770241627808e-07, "loss": 0.4199, "mean_token_accuracy": 0.8623414635658264, "num_tokens": 46230641.0, "step": 1213 }, { "epoch": 0.15443327820887928, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.263583183288574, "learning_rate": 5.142009325985587e-07, "loss": 0.4591, "mean_token_accuracy": 0.8532280921936035, "num_tokens": 46268993.0, "step": 1214 }, { "epoch": 0.15456048848746978, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.238478660583496, "learning_rate": 5.146248410343365e-07, "loss": 0.4924, "mean_token_accuracy": 0.8410637378692627, "num_tokens": 46309310.0, "step": 1215 }, { "epoch": 0.1546876987660603, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.252514362335205, "learning_rate": 5.150487494701145e-07, "loss": 0.4341, "mean_token_accuracy": 0.8582631945610046, "num_tokens": 46352415.0, "step": 1216 }, { "epoch": 0.1548149090446508, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.302628993988037, "learning_rate": 5.154726579058923e-07, "loss": 0.4896, "mean_token_accuracy": 0.8401074409484863, "num_tokens": 46387838.0, "step": 1217 }, { "epoch": 0.1549421193232413, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.209324359893799, "learning_rate": 5.158965663416703e-07, "loss": 0.4847, "mean_token_accuracy": 0.843288779258728, "num_tokens": 46432669.0, "step": 1218 }, { "epoch": 0.15506932960183184, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.245518207550049, "learning_rate": 5.16320474777448e-07, "loss": 0.4405, "mean_token_accuracy": 0.8561527729034424, "num_tokens": 46474383.0, "step": 1219 }, { "epoch": 0.15519653988042234, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.252912521362305, "learning_rate": 5.167443832132259e-07, "loss": 0.4904, "mean_token_accuracy": 0.8405724763870239, "num_tokens": 46514263.0, "step": 1220 }, { "epoch": 0.15532375015901284, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.280857563018799, "learning_rate": 5.171682916490038e-07, "loss": 0.4585, "mean_token_accuracy": 0.8501603007316589, "num_tokens": 46549835.0, "step": 1221 }, { "epoch": 0.15545096043760337, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.334811210632324, "learning_rate": 5.175922000847816e-07, "loss": 0.4454, "mean_token_accuracy": 0.8546537160873413, "num_tokens": 46584811.0, "step": 1222 }, { "epoch": 0.15557817071619387, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.287875175476074, "learning_rate": 5.180161085205595e-07, "loss": 0.467, "mean_token_accuracy": 0.848253071308136, "num_tokens": 46621441.0, "step": 1223 }, { "epoch": 0.15570538099478437, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.269536018371582, "learning_rate": 5.184400169563374e-07, "loss": 0.5198, "mean_token_accuracy": 0.8326276540756226, "num_tokens": 46661572.0, "step": 1224 }, { "epoch": 0.1558325912733749, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.29150390625, "learning_rate": 5.188639253921153e-07, "loss": 0.4948, "mean_token_accuracy": 0.8389350771903992, "num_tokens": 46702157.0, "step": 1225 }, { "epoch": 0.1559598015519654, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.265649795532227, "learning_rate": 5.192878338278932e-07, "loss": 0.4707, "mean_token_accuracy": 0.8474652171134949, "num_tokens": 46739119.0, "step": 1226 }, { "epoch": 0.1560870118305559, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 6.2847580909729, "learning_rate": 5.19711742263671e-07, "loss": 0.5297, "mean_token_accuracy": 0.8304616212844849, "num_tokens": 46783990.0, "step": 1227 }, { "epoch": 0.15621422210914643, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.317330360412598, "learning_rate": 5.201356506994488e-07, "loss": 0.4678, "mean_token_accuracy": 0.842551052570343, "num_tokens": 46816131.0, "step": 1228 }, { "epoch": 0.15634143238773693, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.302933692932129, "learning_rate": 5.205595591352268e-07, "loss": 0.5508, "mean_token_accuracy": 0.8240318298339844, "num_tokens": 46856453.0, "step": 1229 }, { "epoch": 0.15646864266632743, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 6.25083589553833, "learning_rate": 5.209834675710046e-07, "loss": 0.4898, "mean_token_accuracy": 0.8415119647979736, "num_tokens": 46897961.0, "step": 1230 }, { "epoch": 0.15659585294491796, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.33005952835083, "learning_rate": 5.214073760067825e-07, "loss": 0.564, "mean_token_accuracy": 0.8205335140228271, "num_tokens": 46931961.0, "step": 1231 }, { "epoch": 0.15672306322350846, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.310173511505127, "learning_rate": 5.218312844425604e-07, "loss": 0.4847, "mean_token_accuracy": 0.8459162712097168, "num_tokens": 46967822.0, "step": 1232 }, { "epoch": 0.15685027350209896, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.270857810974121, "learning_rate": 5.222551928783383e-07, "loss": 0.4659, "mean_token_accuracy": 0.8481102585792542, "num_tokens": 47004511.0, "step": 1233 }, { "epoch": 0.1569774837806895, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.291309356689453, "learning_rate": 5.226791013141161e-07, "loss": 0.519, "mean_token_accuracy": 0.8352562785148621, "num_tokens": 47047244.0, "step": 1234 }, { "epoch": 0.15710469405928, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.282773971557617, "learning_rate": 5.23103009749894e-07, "loss": 0.4762, "mean_token_accuracy": 0.8462232351303101, "num_tokens": 47088206.0, "step": 1235 }, { "epoch": 0.1572319043378705, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.231889724731445, "learning_rate": 5.235269181856718e-07, "loss": 0.4843, "mean_token_accuracy": 0.8429510593414307, "num_tokens": 47132733.0, "step": 1236 }, { "epoch": 0.15735911461646102, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.330166816711426, "learning_rate": 5.239508266214498e-07, "loss": 0.5423, "mean_token_accuracy": 0.8233712911605835, "num_tokens": 47166840.0, "step": 1237 }, { "epoch": 0.15748632489505152, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.307342529296875, "learning_rate": 5.243747350572276e-07, "loss": 0.5019, "mean_token_accuracy": 0.8383928537368774, "num_tokens": 47203444.0, "step": 1238 }, { "epoch": 0.15761353517364202, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.300205230712891, "learning_rate": 5.247986434930056e-07, "loss": 0.496, "mean_token_accuracy": 0.8429228067398071, "num_tokens": 47240380.0, "step": 1239 }, { "epoch": 0.15774074545223254, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.340014457702637, "learning_rate": 5.252225519287834e-07, "loss": 0.5558, "mean_token_accuracy": 0.822293758392334, "num_tokens": 47277886.0, "step": 1240 }, { "epoch": 0.15786795573082305, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.35829496383667, "learning_rate": 5.256464603645613e-07, "loss": 0.4848, "mean_token_accuracy": 0.8466702699661255, "num_tokens": 47315261.0, "step": 1241 }, { "epoch": 0.15799516600941357, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.291001796722412, "learning_rate": 5.260703688003391e-07, "loss": 0.485, "mean_token_accuracy": 0.8437842726707458, "num_tokens": 47357489.0, "step": 1242 }, { "epoch": 0.15812237628800407, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.344077110290527, "learning_rate": 5.26494277236117e-07, "loss": 0.5005, "mean_token_accuracy": 0.8395110368728638, "num_tokens": 47389851.0, "step": 1243 }, { "epoch": 0.15824958656659457, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 6.34729528427124, "learning_rate": 5.269181856718948e-07, "loss": 0.527, "mean_token_accuracy": 0.8293806314468384, "num_tokens": 47427491.0, "step": 1244 }, { "epoch": 0.1583767968451851, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.319267749786377, "learning_rate": 5.273420941076727e-07, "loss": 0.4675, "mean_token_accuracy": 0.8498809337615967, "num_tokens": 47462790.0, "step": 1245 }, { "epoch": 0.1585040071237756, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.296335697174072, "learning_rate": 5.277660025434506e-07, "loss": 0.4306, "mean_token_accuracy": 0.8583828806877136, "num_tokens": 47502240.0, "step": 1246 }, { "epoch": 0.1586312174023661, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.295074462890625, "learning_rate": 5.281899109792285e-07, "loss": 0.4551, "mean_token_accuracy": 0.8519412279129028, "num_tokens": 47542493.0, "step": 1247 }, { "epoch": 0.15875842768095663, "ewc_loss": 0.005615234375, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 6.289330005645752, "learning_rate": 5.286138194150064e-07, "loss": 0.4918, "mean_token_accuracy": 0.8430755138397217, "num_tokens": 47583378.0, "step": 1248 }, { "epoch": 0.15888563795954713, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.3004326820373535, "learning_rate": 5.290377278507841e-07, "loss": 0.5167, "mean_token_accuracy": 0.8389996886253357, "num_tokens": 47622671.0, "step": 1249 }, { "epoch": 0.15901284823813763, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.3363728523254395, "learning_rate": 5.294616362865621e-07, "loss": 0.4755, "mean_token_accuracy": 0.8494597673416138, "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": 6.319194793701172, "learning_rate": 5.298855447223399e-07, "loss": 0.5461, "mean_token_accuracy": 0.8345506191253662, "num_tokens": 47706143.0, "step": 1251 }, { "epoch": 0.15926726879531866, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.3542985916137695, "learning_rate": 5.303094531581178e-07, "loss": 0.5077, "mean_token_accuracy": 0.8370596766471863, "num_tokens": 47741197.0, "step": 1252 }, { "epoch": 0.15939447907390916, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 6.332573890686035, "learning_rate": 5.307333615938957e-07, "loss": 0.4315, "mean_token_accuracy": 0.858471155166626, "num_tokens": 47779306.0, "step": 1253 }, { "epoch": 0.1595216893524997, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.370169639587402, "learning_rate": 5.311572700296736e-07, "loss": 0.5163, "mean_token_accuracy": 0.833666205406189, "num_tokens": 47811056.0, "step": 1254 }, { "epoch": 0.1596488996310902, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.335071086883545, "learning_rate": 5.315811784654515e-07, "loss": 0.4247, "mean_token_accuracy": 0.8570513725280762, "num_tokens": 47847075.0, "step": 1255 }, { "epoch": 0.1597761099096807, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 6.2713823318481445, "learning_rate": 5.320050869012294e-07, "loss": 0.4821, "mean_token_accuracy": 0.8449423909187317, "num_tokens": 47892639.0, "step": 1256 }, { "epoch": 0.15990332018827122, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 6.3420562744140625, "learning_rate": 5.324289953370071e-07, "loss": 0.5395, "mean_token_accuracy": 0.8289313316345215, "num_tokens": 47933522.0, "step": 1257 }, { "epoch": 0.16003053046686172, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.303774356842041, "learning_rate": 5.328529037727851e-07, "loss": 0.476, "mean_token_accuracy": 0.8443962335586548, "num_tokens": 47977307.0, "step": 1258 }, { "epoch": 0.16015774074545222, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 6.3746018409729, "learning_rate": 5.332768122085629e-07, "loss": 0.4735, "mean_token_accuracy": 0.8448594808578491, "num_tokens": 48015129.0, "step": 1259 }, { "epoch": 0.16028495102404275, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.365937232971191, "learning_rate": 5.337007206443408e-07, "loss": 0.508, "mean_token_accuracy": 0.8333039283752441, "num_tokens": 48051904.0, "step": 1260 }, { "epoch": 0.16041216130263325, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.329385280609131, "learning_rate": 5.341246290801187e-07, "loss": 0.4695, "mean_token_accuracy": 0.8509831428527832, "num_tokens": 48094001.0, "step": 1261 }, { "epoch": 0.16053937158122375, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.308773994445801, "learning_rate": 5.345485375158966e-07, "loss": 0.4622, "mean_token_accuracy": 0.8471687436103821, "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": 6.359987735748291, "learning_rate": 5.349724459516745e-07, "loss": 0.4516, "mean_token_accuracy": 0.8523115515708923, "num_tokens": 48170730.0, "step": 1263 }, { "epoch": 0.16079379213840478, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.3893585205078125, "learning_rate": 5.353963543874522e-07, "loss": 0.523, "mean_token_accuracy": 0.8328193426132202, "num_tokens": 48212640.0, "step": 1264 }, { "epoch": 0.1609210024169953, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.433402061462402, "learning_rate": 5.358202628232301e-07, "loss": 0.5009, "mean_token_accuracy": 0.8392214775085449, "num_tokens": 48243136.0, "step": 1265 }, { "epoch": 0.1610482126955858, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.375523090362549, "learning_rate": 5.36244171259008e-07, "loss": 0.4802, "mean_token_accuracy": 0.8447378873825073, "num_tokens": 48282905.0, "step": 1266 }, { "epoch": 0.1611754229741763, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.341325759887695, "learning_rate": 5.366680796947859e-07, "loss": 0.5185, "mean_token_accuracy": 0.8355814218521118, "num_tokens": 48323325.0, "step": 1267 }, { "epoch": 0.16130263325276684, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 6.4135918617248535, "learning_rate": 5.370919881305637e-07, "loss": 0.4955, "mean_token_accuracy": 0.84345543384552, "num_tokens": 48358400.0, "step": 1268 }, { "epoch": 0.16142984353135734, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.378323078155518, "learning_rate": 5.375158965663417e-07, "loss": 0.4977, "mean_token_accuracy": 0.8411274552345276, "num_tokens": 48393018.0, "step": 1269 }, { "epoch": 0.16155705380994784, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.375587463378906, "learning_rate": 5.379398050021195e-07, "loss": 0.4848, "mean_token_accuracy": 0.8441084623336792, "num_tokens": 48432952.0, "step": 1270 }, { "epoch": 0.16168426408853837, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.419140815734863, "learning_rate": 5.383637134378975e-07, "loss": 0.4727, "mean_token_accuracy": 0.844211995601654, "num_tokens": 48466913.0, "step": 1271 }, { "epoch": 0.16181147436712887, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.377596855163574, "learning_rate": 5.387876218736752e-07, "loss": 0.5532, "mean_token_accuracy": 0.8198039531707764, "num_tokens": 48512658.0, "step": 1272 }, { "epoch": 0.16193868464571937, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 6.367033004760742, "learning_rate": 5.392115303094531e-07, "loss": 0.45, "mean_token_accuracy": 0.8524436354637146, "num_tokens": 48552283.0, "step": 1273 }, { "epoch": 0.1620658949243099, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.411261081695557, "learning_rate": 5.39635438745231e-07, "loss": 0.43, "mean_token_accuracy": 0.8599758148193359, "num_tokens": 48590504.0, "step": 1274 }, { "epoch": 0.1621931052029004, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 6.387231349945068, "learning_rate": 5.400593471810089e-07, "loss": 0.518, "mean_token_accuracy": 0.8338450193405151, "num_tokens": 48629993.0, "step": 1275 }, { "epoch": 0.1623203154814909, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.491640090942383, "learning_rate": 5.404832556167867e-07, "loss": 0.5214, "mean_token_accuracy": 0.8328034281730652, "num_tokens": 48665074.0, "step": 1276 }, { "epoch": 0.16244752576008142, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.400069236755371, "learning_rate": 5.409071640525647e-07, "loss": 0.4794, "mean_token_accuracy": 0.848642110824585, "num_tokens": 48705524.0, "step": 1277 }, { "epoch": 0.16257473603867192, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.398201942443848, "learning_rate": 5.413310724883425e-07, "loss": 0.5263, "mean_token_accuracy": 0.8243870735168457, "num_tokens": 48744030.0, "step": 1278 }, { "epoch": 0.16270194631726242, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.3968071937561035, "learning_rate": 5.417549809241205e-07, "loss": 0.5167, "mean_token_accuracy": 0.8373502492904663, "num_tokens": 48788401.0, "step": 1279 }, { "epoch": 0.16282915659585295, "ewc_loss": 0.005859375, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 6.434013843536377, "learning_rate": 5.421788893598982e-07, "loss": 0.5204, "mean_token_accuracy": 0.8328242301940918, "num_tokens": 48823046.0, "step": 1280 }, { "epoch": 0.16295636687444345, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.405169486999512, "learning_rate": 5.42602797795676e-07, "loss": 0.5265, "mean_token_accuracy": 0.8302347660064697, "num_tokens": 48868293.0, "step": 1281 }, { "epoch": 0.16308357715303395, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.396387577056885, "learning_rate": 5.43026706231454e-07, "loss": 0.5549, "mean_token_accuracy": 0.821363627910614, "num_tokens": 48908775.0, "step": 1282 }, { "epoch": 0.16321078743162448, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.411755561828613, "learning_rate": 5.434506146672319e-07, "loss": 0.5321, "mean_token_accuracy": 0.8294682502746582, "num_tokens": 48957640.0, "step": 1283 }, { "epoch": 0.16333799771021498, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.410734176635742, "learning_rate": 5.438745231030097e-07, "loss": 0.5231, "mean_token_accuracy": 0.829279363155365, "num_tokens": 48993601.0, "step": 1284 }, { "epoch": 0.16346520798880548, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.4375176429748535, "learning_rate": 5.442984315387876e-07, "loss": 0.4299, "mean_token_accuracy": 0.8587199449539185, "num_tokens": 49028411.0, "step": 1285 }, { "epoch": 0.163592418267396, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.4005866050720215, "learning_rate": 5.447223399745655e-07, "loss": 0.4934, "mean_token_accuracy": 0.844018816947937, "num_tokens": 49067449.0, "step": 1286 }, { "epoch": 0.1637196285459865, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.464625358581543, "learning_rate": 5.451462484103433e-07, "loss": 0.4685, "mean_token_accuracy": 0.848447322845459, "num_tokens": 49102650.0, "step": 1287 }, { "epoch": 0.163846838824577, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.45025634765625, "learning_rate": 5.455701568461212e-07, "loss": 0.4921, "mean_token_accuracy": 0.8399196863174438, "num_tokens": 49137110.0, "step": 1288 }, { "epoch": 0.16397404910316754, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 6.479198932647705, "learning_rate": 5.45994065281899e-07, "loss": 0.5637, "mean_token_accuracy": 0.8179064393043518, "num_tokens": 49177765.0, "step": 1289 }, { "epoch": 0.16410125938175804, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.465250015258789, "learning_rate": 5.46417973717677e-07, "loss": 0.5043, "mean_token_accuracy": 0.8342435956001282, "num_tokens": 49207860.0, "step": 1290 }, { "epoch": 0.16422846966034857, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 6.471492290496826, "learning_rate": 5.468418821534548e-07, "loss": 0.4371, "mean_token_accuracy": 0.8563075065612793, "num_tokens": 49239994.0, "step": 1291 }, { "epoch": 0.16435567993893907, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 6.440876483917236, "learning_rate": 5.472657905892327e-07, "loss": 0.452, "mean_token_accuracy": 0.8547141551971436, "num_tokens": 49273464.0, "step": 1292 }, { "epoch": 0.16448289021752957, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 6.455560207366943, "learning_rate": 5.476896990250106e-07, "loss": 0.5654, "mean_token_accuracy": 0.8203819990158081, "num_tokens": 49313144.0, "step": 1293 }, { "epoch": 0.1646101004961201, "ewc_loss": 0.006011962890625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.428671836853027, "learning_rate": 5.481136074607885e-07, "loss": 0.452, "mean_token_accuracy": 0.8494906425476074, "num_tokens": 49351392.0, "step": 1294 }, { "epoch": 0.1647373107747106, "ewc_loss": 0.005889892578125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 6.388892650604248, "learning_rate": 5.485375158965663e-07, "loss": 0.4541, "mean_token_accuracy": 0.8529806137084961, "num_tokens": 49396726.0, "step": 1295 }, { "epoch": 0.1648645210533011, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 6.42588996887207, "learning_rate": 5.489614243323442e-07, "loss": 0.4879, "mean_token_accuracy": 0.8413177132606506, "num_tokens": 49435856.0, "step": 1296 }, { "epoch": 0.16499173133189163, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 6.427424430847168, "learning_rate": 5.49385332768122e-07, "loss": 0.4286, "mean_token_accuracy": 0.8576158285140991, "num_tokens": 49473695.0, "step": 1297 }, { "epoch": 0.16511894161048213, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.486908912658691, "learning_rate": 5.498092412039e-07, "loss": 0.5544, "mean_token_accuracy": 0.8272638916969299, "num_tokens": 49509165.0, "step": 1298 }, { "epoch": 0.16524615188907263, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.398029327392578, "learning_rate": 5.502331496396778e-07, "loss": 0.4589, "mean_token_accuracy": 0.8501714468002319, "num_tokens": 49553790.0, "step": 1299 }, { "epoch": 0.16537336216766316, "ewc_loss": 0.006011962890625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.45740270614624, "learning_rate": 5.506570580754557e-07, "loss": 0.4826, "mean_token_accuracy": 0.8430958390235901, "num_tokens": 49593124.0, "step": 1300 }, { "epoch": 0.16550057244625366, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.48239803314209, "learning_rate": 5.510809665112336e-07, "loss": 0.4698, "mean_token_accuracy": 0.8504914045333862, "num_tokens": 49632016.0, "step": 1301 }, { "epoch": 0.16562778272484416, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.472379684448242, "learning_rate": 5.515048749470113e-07, "loss": 0.4851, "mean_token_accuracy": 0.8444947004318237, "num_tokens": 49670689.0, "step": 1302 }, { "epoch": 0.1657549930034347, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.485968112945557, "learning_rate": 5.519287833827893e-07, "loss": 0.4708, "mean_token_accuracy": 0.8467879295349121, "num_tokens": 49709230.0, "step": 1303 }, { "epoch": 0.1658822032820252, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.502851963043213, "learning_rate": 5.523526918185671e-07, "loss": 0.5265, "mean_token_accuracy": 0.8298327326774597, "num_tokens": 49741648.0, "step": 1304 }, { "epoch": 0.1660094135606157, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.507424831390381, "learning_rate": 5.52776600254345e-07, "loss": 0.4453, "mean_token_accuracy": 0.8516424298286438, "num_tokens": 49773906.0, "step": 1305 }, { "epoch": 0.16613662383920622, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 6.447624683380127, "learning_rate": 5.532005086901229e-07, "loss": 0.5045, "mean_token_accuracy": 0.8420161604881287, "num_tokens": 49812645.0, "step": 1306 }, { "epoch": 0.16626383411779672, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.426093101501465, "learning_rate": 5.536244171259008e-07, "loss": 0.4736, "mean_token_accuracy": 0.8475788235664368, "num_tokens": 49854500.0, "step": 1307 }, { "epoch": 0.16639104439638722, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 6.564011573791504, "learning_rate": 5.540483255616786e-07, "loss": 0.5669, "mean_token_accuracy": 0.8201782703399658, "num_tokens": 49889408.0, "step": 1308 }, { "epoch": 0.16651825467497774, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.522188663482666, "learning_rate": 5.544722339974566e-07, "loss": 0.4899, "mean_token_accuracy": 0.8424030542373657, "num_tokens": 49923581.0, "step": 1309 }, { "epoch": 0.16664546495356825, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.50075626373291, "learning_rate": 5.548961424332343e-07, "loss": 0.4787, "mean_token_accuracy": 0.8424363136291504, "num_tokens": 49959433.0, "step": 1310 }, { "epoch": 0.16677267523215875, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.461362361907959, "learning_rate": 5.553200508690123e-07, "loss": 0.4466, "mean_token_accuracy": 0.8558509349822998, "num_tokens": 50000307.0, "step": 1311 }, { "epoch": 0.16689988551074927, "ewc_loss": 0.006103515625, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 6.504172325134277, "learning_rate": 5.557439593047901e-07, "loss": 0.5401, "mean_token_accuracy": 0.8255537748336792, "num_tokens": 50035587.0, "step": 1312 }, { "epoch": 0.16702709578933977, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.464712619781494, "learning_rate": 5.56167867740568e-07, "loss": 0.4783, "mean_token_accuracy": 0.844896137714386, "num_tokens": 50079425.0, "step": 1313 }, { "epoch": 0.16715430606793028, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.5097551345825195, "learning_rate": 5.565917761763459e-07, "loss": 0.4613, "mean_token_accuracy": 0.8509621620178223, "num_tokens": 50110674.0, "step": 1314 }, { "epoch": 0.1672815163465208, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.520308971405029, "learning_rate": 5.570156846121238e-07, "loss": 0.4998, "mean_token_accuracy": 0.8373076915740967, "num_tokens": 50149033.0, "step": 1315 }, { "epoch": 0.1674087266251113, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.52210807800293, "learning_rate": 5.574395930479016e-07, "loss": 0.5201, "mean_token_accuracy": 0.8331476449966431, "num_tokens": 50185728.0, "step": 1316 }, { "epoch": 0.16753593690370183, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.525023937225342, "learning_rate": 5.578635014836796e-07, "loss": 0.4702, "mean_token_accuracy": 0.8480266332626343, "num_tokens": 50219671.0, "step": 1317 }, { "epoch": 0.16766314718229233, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.5469841957092285, "learning_rate": 5.582874099194573e-07, "loss": 0.5293, "mean_token_accuracy": 0.8339332342147827, "num_tokens": 50251989.0, "step": 1318 }, { "epoch": 0.16779035746088283, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.485662460327148, "learning_rate": 5.587113183552353e-07, "loss": 0.5145, "mean_token_accuracy": 0.8350173234939575, "num_tokens": 50291079.0, "step": 1319 }, { "epoch": 0.16791756773947336, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 6.490618705749512, "learning_rate": 5.591352267910131e-07, "loss": 0.4709, "mean_token_accuracy": 0.847405195236206, "num_tokens": 50332262.0, "step": 1320 }, { "epoch": 0.16804477801806386, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.550690174102783, "learning_rate": 5.59559135226791e-07, "loss": 0.5007, "mean_token_accuracy": 0.841499924659729, "num_tokens": 50371278.0, "step": 1321 }, { "epoch": 0.16817198829665436, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.552891254425049, "learning_rate": 5.599830436625689e-07, "loss": 0.4781, "mean_token_accuracy": 0.8462998867034912, "num_tokens": 50409463.0, "step": 1322 }, { "epoch": 0.1682991985752449, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.531866073608398, "learning_rate": 5.604069520983468e-07, "loss": 0.4725, "mean_token_accuracy": 0.8483310341835022, "num_tokens": 50446745.0, "step": 1323 }, { "epoch": 0.1684264088538354, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.553776741027832, "learning_rate": 5.608308605341246e-07, "loss": 0.4639, "mean_token_accuracy": 0.8484967947006226, "num_tokens": 50479725.0, "step": 1324 }, { "epoch": 0.1685536191324259, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.571784019470215, "learning_rate": 5.612547689699024e-07, "loss": 0.5266, "mean_token_accuracy": 0.8317074775695801, "num_tokens": 50514734.0, "step": 1325 }, { "epoch": 0.16868082941101642, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.491825580596924, "learning_rate": 5.616786774056803e-07, "loss": 0.505, "mean_token_accuracy": 0.8402783870697021, "num_tokens": 50555165.0, "step": 1326 }, { "epoch": 0.16880803968960692, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.486754894256592, "learning_rate": 5.621025858414582e-07, "loss": 0.5059, "mean_token_accuracy": 0.8365057110786438, "num_tokens": 50598251.0, "step": 1327 }, { "epoch": 0.16893524996819742, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.534408092498779, "learning_rate": 5.625264942772361e-07, "loss": 0.4978, "mean_token_accuracy": 0.8419321179389954, "num_tokens": 50638423.0, "step": 1328 }, { "epoch": 0.16906246024678795, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.4927215576171875, "learning_rate": 5.629504027130139e-07, "loss": 0.5556, "mean_token_accuracy": 0.8314177393913269, "num_tokens": 50678226.0, "step": 1329 }, { "epoch": 0.16918967052537845, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.526150703430176, "learning_rate": 5.633743111487919e-07, "loss": 0.5563, "mean_token_accuracy": 0.8199293613433838, "num_tokens": 50717542.0, "step": 1330 }, { "epoch": 0.16931688080396895, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.536275386810303, "learning_rate": 5.637982195845697e-07, "loss": 0.4369, "mean_token_accuracy": 0.858281672000885, "num_tokens": 50752914.0, "step": 1331 }, { "epoch": 0.16944409108255948, "ewc_loss": 0.00616455078125, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 6.511160850524902, "learning_rate": 5.642221280203476e-07, "loss": 0.4358, "mean_token_accuracy": 0.859422504901886, "num_tokens": 50789276.0, "step": 1332 }, { "epoch": 0.16957130136114998, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.627142906188965, "learning_rate": 5.646460364561254e-07, "loss": 0.4564, "mean_token_accuracy": 0.8497723340988159, "num_tokens": 50820759.0, "step": 1333 }, { "epoch": 0.16969851163974048, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.541615009307861, "learning_rate": 5.650699448919033e-07, "loss": 0.5471, "mean_token_accuracy": 0.8255308866500854, "num_tokens": 50863241.0, "step": 1334 }, { "epoch": 0.169825721918331, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.537369251251221, "learning_rate": 5.654938533276812e-07, "loss": 0.4858, "mean_token_accuracy": 0.8418805599212646, "num_tokens": 50904296.0, "step": 1335 }, { "epoch": 0.1699529321969215, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.539883613586426, "learning_rate": 5.659177617634591e-07, "loss": 0.4726, "mean_token_accuracy": 0.8477615714073181, "num_tokens": 50942180.0, "step": 1336 }, { "epoch": 0.170080142475512, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.533646106719971, "learning_rate": 5.663416701992369e-07, "loss": 0.4984, "mean_token_accuracy": 0.8410017490386963, "num_tokens": 50980287.0, "step": 1337 }, { "epoch": 0.17020735275410254, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.586490631103516, "learning_rate": 5.667655786350149e-07, "loss": 0.5606, "mean_token_accuracy": 0.8148807287216187, "num_tokens": 51014218.0, "step": 1338 }, { "epoch": 0.17033456303269304, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.562233924865723, "learning_rate": 5.671894870707927e-07, "loss": 0.4784, "mean_token_accuracy": 0.848331093788147, "num_tokens": 51055915.0, "step": 1339 }, { "epoch": 0.17046177331128357, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.535707473754883, "learning_rate": 5.676133955065705e-07, "loss": 0.4226, "mean_token_accuracy": 0.8595025539398193, "num_tokens": 51090684.0, "step": 1340 }, { "epoch": 0.17058898358987407, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.590828895568848, "learning_rate": 5.680373039423484e-07, "loss": 0.4686, "mean_token_accuracy": 0.8475354909896851, "num_tokens": 51130156.0, "step": 1341 }, { "epoch": 0.17071619386846457, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.589629650115967, "learning_rate": 5.684612123781263e-07, "loss": 0.509, "mean_token_accuracy": 0.8348555564880371, "num_tokens": 51171282.0, "step": 1342 }, { "epoch": 0.1708434041470551, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.563384532928467, "learning_rate": 5.688851208139042e-07, "loss": 0.5245, "mean_token_accuracy": 0.8299927711486816, "num_tokens": 51210616.0, "step": 1343 }, { "epoch": 0.1709706144256456, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.582949638366699, "learning_rate": 5.69309029249682e-07, "loss": 0.5106, "mean_token_accuracy": 0.8412032127380371, "num_tokens": 51253357.0, "step": 1344 }, { "epoch": 0.1710978247042361, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.57448673248291, "learning_rate": 5.697329376854599e-07, "loss": 0.4909, "mean_token_accuracy": 0.8405008316040039, "num_tokens": 51292671.0, "step": 1345 }, { "epoch": 0.17122503498282662, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.602436542510986, "learning_rate": 5.701568461212378e-07, "loss": 0.4302, "mean_token_accuracy": 0.8601958751678467, "num_tokens": 51329145.0, "step": 1346 }, { "epoch": 0.17135224526141712, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.583041191101074, "learning_rate": 5.705807545570157e-07, "loss": 0.44, "mean_token_accuracy": 0.855916440486908, "num_tokens": 51371265.0, "step": 1347 }, { "epoch": 0.17147945554000762, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.616291522979736, "learning_rate": 5.710046629927934e-07, "loss": 0.4762, "mean_token_accuracy": 0.8444607853889465, "num_tokens": 51410262.0, "step": 1348 }, { "epoch": 0.17160666581859815, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.5612263679504395, "learning_rate": 5.714285714285714e-07, "loss": 0.4476, "mean_token_accuracy": 0.8549555540084839, "num_tokens": 51447796.0, "step": 1349 }, { "epoch": 0.17173387609718865, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.687219142913818, "learning_rate": 5.718524798643492e-07, "loss": 0.5113, "mean_token_accuracy": 0.8378782272338867, "num_tokens": 51482677.0, "step": 1350 }, { "epoch": 0.17186108637577915, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.606564044952393, "learning_rate": 5.722763883001272e-07, "loss": 0.5121, "mean_token_accuracy": 0.8366941213607788, "num_tokens": 51524111.0, "step": 1351 }, { "epoch": 0.17198829665436968, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 6.55417013168335, "learning_rate": 5.72700296735905e-07, "loss": 0.4845, "mean_token_accuracy": 0.8423328995704651, "num_tokens": 51561008.0, "step": 1352 }, { "epoch": 0.17211550693296018, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.609822750091553, "learning_rate": 5.731242051716829e-07, "loss": 0.5388, "mean_token_accuracy": 0.824311375617981, "num_tokens": 51603091.0, "step": 1353 }, { "epoch": 0.17224271721155068, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.5908684730529785, "learning_rate": 5.735481136074608e-07, "loss": 0.4542, "mean_token_accuracy": 0.8505649566650391, "num_tokens": 51637789.0, "step": 1354 }, { "epoch": 0.1723699274901412, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.607344150543213, "learning_rate": 5.739720220432386e-07, "loss": 0.4586, "mean_token_accuracy": 0.8485606908798218, "num_tokens": 51674207.0, "step": 1355 }, { "epoch": 0.1724971377687317, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.608935832977295, "learning_rate": 5.743959304790164e-07, "loss": 0.5545, "mean_token_accuracy": 0.8241017460823059, "num_tokens": 51715109.0, "step": 1356 }, { "epoch": 0.1726243480473222, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.630552291870117, "learning_rate": 5.748198389147944e-07, "loss": 0.5012, "mean_token_accuracy": 0.8410964012145996, "num_tokens": 51753373.0, "step": 1357 }, { "epoch": 0.17275155832591274, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.5885329246521, "learning_rate": 5.752437473505722e-07, "loss": 0.5386, "mean_token_accuracy": 0.8307453989982605, "num_tokens": 51795100.0, "step": 1358 }, { "epoch": 0.17287876860450324, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.611730098724365, "learning_rate": 5.756676557863502e-07, "loss": 0.5131, "mean_token_accuracy": 0.8341455459594727, "num_tokens": 51831840.0, "step": 1359 }, { "epoch": 0.17300597888309374, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.5916056632995605, "learning_rate": 5.76091564222128e-07, "loss": 0.4652, "mean_token_accuracy": 0.8496730327606201, "num_tokens": 51870227.0, "step": 1360 }, { "epoch": 0.17313318916168427, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 1.2278556823730469e-05, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 6.637453079223633, "learning_rate": 5.765154726579059e-07, "loss": 0.4904, "mean_token_accuracy": 0.8425226807594299, "num_tokens": 51908251.0, "step": 1361 }, { "epoch": 0.17326039944027477, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 1.2278556823730469e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.618388652801514, "learning_rate": 5.769393810936838e-07, "loss": 0.4803, "mean_token_accuracy": 0.8451428413391113, "num_tokens": 51944531.0, "step": 1362 }, { "epoch": 0.17338760971886527, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.2278556823730469e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.607318878173828, "learning_rate": 5.773632895294616e-07, "loss": 0.5334, "mean_token_accuracy": 0.8278055787086487, "num_tokens": 51985742.0, "step": 1363 }, { "epoch": 0.1735148199974558, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 6.606207847595215, "learning_rate": 5.777871979652394e-07, "loss": 0.4712, "mean_token_accuracy": 0.8475229740142822, "num_tokens": 52022036.0, "step": 1364 }, { "epoch": 0.1736420302760463, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.614044189453125, "learning_rate": 5.782111064010173e-07, "loss": 0.4944, "mean_token_accuracy": 0.8396926522254944, "num_tokens": 52066077.0, "step": 1365 }, { "epoch": 0.17376924055463683, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.563909530639648, "learning_rate": 5.786350148367952e-07, "loss": 0.4856, "mean_token_accuracy": 0.8430027961730957, "num_tokens": 52111280.0, "step": 1366 }, { "epoch": 0.17389645083322733, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.594635486602783, "learning_rate": 5.790589232725731e-07, "loss": 0.4839, "mean_token_accuracy": 0.841480016708374, "num_tokens": 52153148.0, "step": 1367 }, { "epoch": 0.17402366111181783, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.63790225982666, "learning_rate": 5.79482831708351e-07, "loss": 0.4689, "mean_token_accuracy": 0.8476389646530151, "num_tokens": 52186722.0, "step": 1368 }, { "epoch": 0.17415087139040836, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.65554666519165, "learning_rate": 5.799067401441288e-07, "loss": 0.4786, "mean_token_accuracy": 0.8431150317192078, "num_tokens": 52220837.0, "step": 1369 }, { "epoch": 0.17427808166899886, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.644742965698242, "learning_rate": 5.803306485799068e-07, "loss": 0.4633, "mean_token_accuracy": 0.8510348200798035, "num_tokens": 52257049.0, "step": 1370 }, { "epoch": 0.17440529194758936, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.591344356536865, "learning_rate": 5.807545570156845e-07, "loss": 0.4407, "mean_token_accuracy": 0.8548077940940857, "num_tokens": 52297632.0, "step": 1371 }, { "epoch": 0.1745325022261799, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 6.7207841873168945, "learning_rate": 5.811784654514624e-07, "loss": 0.5389, "mean_token_accuracy": 0.8273686766624451, "num_tokens": 52331008.0, "step": 1372 }, { "epoch": 0.1746597125047704, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.624546051025391, "learning_rate": 5.816023738872403e-07, "loss": 0.4708, "mean_token_accuracy": 0.8490444421768188, "num_tokens": 52369420.0, "step": 1373 }, { "epoch": 0.1747869227833609, "ewc_loss": 0.00634765625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 6.610655784606934, "learning_rate": 5.820262823230182e-07, "loss": 0.5005, "mean_token_accuracy": 0.8387206196784973, "num_tokens": 52411759.0, "step": 1374 }, { "epoch": 0.17491413306195142, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.708961486816406, "learning_rate": 5.824501907587961e-07, "loss": 0.4613, "mean_token_accuracy": 0.850111722946167, "num_tokens": 52446459.0, "step": 1375 }, { "epoch": 0.17504134334054192, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.666767120361328, "learning_rate": 5.82874099194574e-07, "loss": 0.5448, "mean_token_accuracy": 0.8238340616226196, "num_tokens": 52484615.0, "step": 1376 }, { "epoch": 0.17516855361913242, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.648885726928711, "learning_rate": 5.832980076303518e-07, "loss": 0.5556, "mean_token_accuracy": 0.8213582634925842, "num_tokens": 52525195.0, "step": 1377 }, { "epoch": 0.17529576389772294, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.704251766204834, "learning_rate": 5.837219160661297e-07, "loss": 0.4398, "mean_token_accuracy": 0.8560754060745239, "num_tokens": 52558783.0, "step": 1378 }, { "epoch": 0.17542297417631345, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.6576619148254395, "learning_rate": 5.841458245019075e-07, "loss": 0.4353, "mean_token_accuracy": 0.8565213680267334, "num_tokens": 52599176.0, "step": 1379 }, { "epoch": 0.17555018445490395, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.716465950012207, "learning_rate": 5.845697329376855e-07, "loss": 0.4914, "mean_token_accuracy": 0.84503173828125, "num_tokens": 52630865.0, "step": 1380 }, { "epoch": 0.17567739473349447, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.6924848556518555, "learning_rate": 5.849936413734633e-07, "loss": 0.5428, "mean_token_accuracy": 0.8254973888397217, "num_tokens": 52671702.0, "step": 1381 }, { "epoch": 0.17580460501208497, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.646162033081055, "learning_rate": 5.854175498092412e-07, "loss": 0.4478, "mean_token_accuracy": 0.8551150560379028, "num_tokens": 52712288.0, "step": 1382 }, { "epoch": 0.17593181529067548, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.658777236938477, "learning_rate": 5.858414582450191e-07, "loss": 0.4902, "mean_token_accuracy": 0.842768669128418, "num_tokens": 52756569.0, "step": 1383 }, { "epoch": 0.176059025569266, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.679561138153076, "learning_rate": 5.86265366680797e-07, "loss": 0.5127, "mean_token_accuracy": 0.8367277383804321, "num_tokens": 52796195.0, "step": 1384 }, { "epoch": 0.1761862358478565, "ewc_loss": 0.006439208984375, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.658205032348633, "learning_rate": 5.866892751165748e-07, "loss": 0.4837, "mean_token_accuracy": 0.8454458713531494, "num_tokens": 52839687.0, "step": 1385 }, { "epoch": 0.176313446126447, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.701170921325684, "learning_rate": 5.871131835523526e-07, "loss": 0.5385, "mean_token_accuracy": 0.82807457447052, "num_tokens": 52875660.0, "step": 1386 }, { "epoch": 0.17644065640503753, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.692770957946777, "learning_rate": 5.875370919881305e-07, "loss": 0.4947, "mean_token_accuracy": 0.838823139667511, "num_tokens": 52921220.0, "step": 1387 }, { "epoch": 0.17656786668362803, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.696761131286621, "learning_rate": 5.879610004239084e-07, "loss": 0.548, "mean_token_accuracy": 0.828084409236908, "num_tokens": 52960042.0, "step": 1388 }, { "epoch": 0.17669507696221853, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 6.70854377746582, "learning_rate": 5.883849088596863e-07, "loss": 0.5322, "mean_token_accuracy": 0.8310658931732178, "num_tokens": 53002817.0, "step": 1389 }, { "epoch": 0.17682228724080906, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2576580047607422e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.7073493003845215, "learning_rate": 5.888088172954641e-07, "loss": 0.5055, "mean_token_accuracy": 0.8383580446243286, "num_tokens": 53037751.0, "step": 1390 }, { "epoch": 0.17694949751939956, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.988428592681885, "learning_rate": 5.892327257312421e-07, "loss": 0.4191, "mean_token_accuracy": 0.8650180697441101, "num_tokens": 53074286.0, "step": 1391 }, { "epoch": 0.1770767077979901, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.6929216384887695, "learning_rate": 5.896566341670199e-07, "loss": 0.4402, "mean_token_accuracy": 0.8519399166107178, "num_tokens": 53113525.0, "step": 1392 }, { "epoch": 0.1772039180765806, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 6.643890857696533, "learning_rate": 5.900805426027977e-07, "loss": 0.4833, "mean_token_accuracy": 0.8431232571601868, "num_tokens": 53150137.0, "step": 1393 }, { "epoch": 0.1773311283551711, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.647455215454102, "learning_rate": 5.905044510385756e-07, "loss": 0.4363, "mean_token_accuracy": 0.8575769066810608, "num_tokens": 53189519.0, "step": 1394 }, { "epoch": 0.17745833863376162, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 6.687368869781494, "learning_rate": 5.909283594743535e-07, "loss": 0.547, "mean_token_accuracy": 0.8247724771499634, "num_tokens": 53227452.0, "step": 1395 }, { "epoch": 0.17758554891235212, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.718006610870361, "learning_rate": 5.913522679101314e-07, "loss": 0.4791, "mean_token_accuracy": 0.8412714600563049, "num_tokens": 53265875.0, "step": 1396 }, { "epoch": 0.17771275919094262, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.676797389984131, "learning_rate": 5.917761763459093e-07, "loss": 0.5392, "mean_token_accuracy": 0.8267426490783691, "num_tokens": 53305477.0, "step": 1397 }, { "epoch": 0.17783996946953315, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.732753753662109, "learning_rate": 5.922000847816871e-07, "loss": 0.4592, "mean_token_accuracy": 0.8523365259170532, "num_tokens": 53342800.0, "step": 1398 }, { "epoch": 0.17796717974812365, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.711300849914551, "learning_rate": 5.926239932174651e-07, "loss": 0.5237, "mean_token_accuracy": 0.8325557708740234, "num_tokens": 53387476.0, "step": 1399 }, { "epoch": 0.17809439002671415, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.673966407775879, "learning_rate": 5.930479016532429e-07, "loss": 0.4805, "mean_token_accuracy": 0.8435002565383911, "num_tokens": 53424628.0, "step": 1400 }, { "epoch": 0.17822160030530468, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.704697132110596, "learning_rate": 5.934718100890207e-07, "loss": 0.4618, "mean_token_accuracy": 0.8480627536773682, "num_tokens": 53464459.0, "step": 1401 }, { "epoch": 0.17834881058389518, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.755587100982666, "learning_rate": 5.938957185247986e-07, "loss": 0.5098, "mean_token_accuracy": 0.8333993554115295, "num_tokens": 53501721.0, "step": 1402 }, { "epoch": 0.17847602086248568, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.740932941436768, "learning_rate": 5.943196269605765e-07, "loss": 0.4712, "mean_token_accuracy": 0.8432870507240295, "num_tokens": 53536607.0, "step": 1403 }, { "epoch": 0.1786032311410762, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.763407230377197, "learning_rate": 5.947435353963544e-07, "loss": 0.5219, "mean_token_accuracy": 0.8359178900718689, "num_tokens": 53569809.0, "step": 1404 }, { "epoch": 0.1787304414196667, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.696069240570068, "learning_rate": 5.951674438321323e-07, "loss": 0.4984, "mean_token_accuracy": 0.8398571014404297, "num_tokens": 53604423.0, "step": 1405 }, { "epoch": 0.1788576516982572, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.751913070678711, "learning_rate": 5.955913522679101e-07, "loss": 0.4941, "mean_token_accuracy": 0.8372278213500977, "num_tokens": 53642145.0, "step": 1406 }, { "epoch": 0.17898486197684774, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.703505992889404, "learning_rate": 5.96015260703688e-07, "loss": 0.5694, "mean_token_accuracy": 0.8161524534225464, "num_tokens": 53682998.0, "step": 1407 }, { "epoch": 0.17911207225543824, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 6.7121968269348145, "learning_rate": 5.964391691394659e-07, "loss": 0.5024, "mean_token_accuracy": 0.840008556842804, "num_tokens": 53729764.0, "step": 1408 }, { "epoch": 0.17923928253402874, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.811584949493408, "learning_rate": 5.968630775752436e-07, "loss": 0.5353, "mean_token_accuracy": 0.8292477130889893, "num_tokens": 53764522.0, "step": 1409 }, { "epoch": 0.17936649281261927, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.785684585571289, "learning_rate": 5.972869860110216e-07, "loss": 0.5178, "mean_token_accuracy": 0.8303827047348022, "num_tokens": 53799791.0, "step": 1410 }, { "epoch": 0.17949370309120977, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.760732173919678, "learning_rate": 5.977108944467994e-07, "loss": 0.5563, "mean_token_accuracy": 0.8257066607475281, "num_tokens": 53840181.0, "step": 1411 }, { "epoch": 0.17962091336980027, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.719676971435547, "learning_rate": 5.981348028825774e-07, "loss": 0.4626, "mean_token_accuracy": 0.8523883819580078, "num_tokens": 53882813.0, "step": 1412 }, { "epoch": 0.1797481236483908, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.812598705291748, "learning_rate": 5.985587113183552e-07, "loss": 0.5109, "mean_token_accuracy": 0.8364235162734985, "num_tokens": 53922198.0, "step": 1413 }, { "epoch": 0.1798753339269813, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.742204189300537, "learning_rate": 5.989826197541331e-07, "loss": 0.5059, "mean_token_accuracy": 0.83802330493927, "num_tokens": 53959390.0, "step": 1414 }, { "epoch": 0.18000254420557182, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.754868984222412, "learning_rate": 5.99406528189911e-07, "loss": 0.5335, "mean_token_accuracy": 0.8341553211212158, "num_tokens": 54005707.0, "step": 1415 }, { "epoch": 0.18012975448416232, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.7430419921875, "learning_rate": 5.998304366256888e-07, "loss": 0.4976, "mean_token_accuracy": 0.8388912677764893, "num_tokens": 54047808.0, "step": 1416 }, { "epoch": 0.18025696476275282, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.777463912963867, "learning_rate": 6.002543450614666e-07, "loss": 0.4225, "mean_token_accuracy": 0.8654230237007141, "num_tokens": 54082929.0, "step": 1417 }, { "epoch": 0.18038417504134335, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.768519878387451, "learning_rate": 6.006782534972446e-07, "loss": 0.483, "mean_token_accuracy": 0.8442460894584656, "num_tokens": 54120199.0, "step": 1418 }, { "epoch": 0.18051138531993385, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.813265323638916, "learning_rate": 6.011021619330224e-07, "loss": 0.519, "mean_token_accuracy": 0.8317649364471436, "num_tokens": 54158629.0, "step": 1419 }, { "epoch": 0.18063859559852435, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.783614158630371, "learning_rate": 6.015260703688004e-07, "loss": 0.4982, "mean_token_accuracy": 0.8365684151649475, "num_tokens": 54194772.0, "step": 1420 }, { "epoch": 0.18076580587711488, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.80639123916626, "learning_rate": 6.019499788045782e-07, "loss": 0.5062, "mean_token_accuracy": 0.8376586437225342, "num_tokens": 54229825.0, "step": 1421 }, { "epoch": 0.18089301615570538, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 6.785549640655518, "learning_rate": 6.023738872403561e-07, "loss": 0.4623, "mean_token_accuracy": 0.8481830358505249, "num_tokens": 54260051.0, "step": 1422 }, { "epoch": 0.18102022643429588, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.771795749664307, "learning_rate": 6.02797795676134e-07, "loss": 0.4891, "mean_token_accuracy": 0.8457766771316528, "num_tokens": 54297865.0, "step": 1423 }, { "epoch": 0.1811474367128864, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.744872093200684, "learning_rate": 6.032217041119118e-07, "loss": 0.5036, "mean_token_accuracy": 0.8332654237747192, "num_tokens": 54334618.0, "step": 1424 }, { "epoch": 0.1812746469914769, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.787081718444824, "learning_rate": 6.036456125476896e-07, "loss": 0.5398, "mean_token_accuracy": 0.8253757953643799, "num_tokens": 54369761.0, "step": 1425 }, { "epoch": 0.1814018572700674, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.752249240875244, "learning_rate": 6.040695209834675e-07, "loss": 0.4398, "mean_token_accuracy": 0.8543515205383301, "num_tokens": 54411549.0, "step": 1426 }, { "epoch": 0.18152906754865794, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.753332614898682, "learning_rate": 6.044934294192454e-07, "loss": 0.525, "mean_token_accuracy": 0.8324090838432312, "num_tokens": 54449137.0, "step": 1427 }, { "epoch": 0.18165627782724844, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.718960762023926, "learning_rate": 6.049173378550233e-07, "loss": 0.4799, "mean_token_accuracy": 0.8470079898834229, "num_tokens": 54494604.0, "step": 1428 }, { "epoch": 0.18178348810583894, "ewc_loss": 0.006591796875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.753048896789551, "learning_rate": 6.053412462908012e-07, "loss": 0.4603, "mean_token_accuracy": 0.8484853506088257, "num_tokens": 54533533.0, "step": 1429 }, { "epoch": 0.18191069838442947, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.788109302520752, "learning_rate": 6.05765154726579e-07, "loss": 0.5328, "mean_token_accuracy": 0.8306950926780701, "num_tokens": 54575211.0, "step": 1430 }, { "epoch": 0.18203790866301997, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.732775688171387, "learning_rate": 6.061890631623569e-07, "loss": 0.4382, "mean_token_accuracy": 0.8566931486129761, "num_tokens": 54618308.0, "step": 1431 }, { "epoch": 0.18216511894161047, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 6.7843217849731445, "learning_rate": 6.066129715981347e-07, "loss": 0.4768, "mean_token_accuracy": 0.8454768657684326, "num_tokens": 54660260.0, "step": 1432 }, { "epoch": 0.182292329220201, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.869232654571533, "learning_rate": 6.070368800339126e-07, "loss": 0.4482, "mean_token_accuracy": 0.8534826040267944, "num_tokens": 54692056.0, "step": 1433 }, { "epoch": 0.1824195394987915, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.802524566650391, "learning_rate": 6.074607884696905e-07, "loss": 0.4696, "mean_token_accuracy": 0.8457661271095276, "num_tokens": 54729252.0, "step": 1434 }, { "epoch": 0.182546749777382, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 6.829219818115234, "learning_rate": 6.078846969054684e-07, "loss": 0.4926, "mean_token_accuracy": 0.8394517302513123, "num_tokens": 54763422.0, "step": 1435 }, { "epoch": 0.18267396005597253, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.78093957901001, "learning_rate": 6.083086053412463e-07, "loss": 0.469, "mean_token_accuracy": 0.8480437994003296, "num_tokens": 54803494.0, "step": 1436 }, { "epoch": 0.18280117033456303, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.871450901031494, "learning_rate": 6.087325137770242e-07, "loss": 0.4575, "mean_token_accuracy": 0.851096510887146, "num_tokens": 54839349.0, "step": 1437 }, { "epoch": 0.18292838061315353, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.806023597717285, "learning_rate": 6.09156422212802e-07, "loss": 0.531, "mean_token_accuracy": 0.8328871130943298, "num_tokens": 54881941.0, "step": 1438 }, { "epoch": 0.18305559089174406, "ewc_loss": 0.0067138671875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 6.819374084472656, "learning_rate": 6.095803306485799e-07, "loss": 0.4815, "mean_token_accuracy": 0.8420470952987671, "num_tokens": 54922927.0, "step": 1439 }, { "epoch": 0.18318280117033456, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.8217549324035645, "learning_rate": 6.100042390843577e-07, "loss": 0.509, "mean_token_accuracy": 0.8354673385620117, "num_tokens": 54963885.0, "step": 1440 }, { "epoch": 0.1833100114489251, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.841897487640381, "learning_rate": 6.104281475201356e-07, "loss": 0.4591, "mean_token_accuracy": 0.8496294021606445, "num_tokens": 55001284.0, "step": 1441 }, { "epoch": 0.1834372217275156, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.850818157196045, "learning_rate": 6.108520559559135e-07, "loss": 0.5188, "mean_token_accuracy": 0.8329474329948425, "num_tokens": 55034813.0, "step": 1442 }, { "epoch": 0.1835644320061061, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.87042236328125, "learning_rate": 6.112759643916914e-07, "loss": 0.4248, "mean_token_accuracy": 0.8610188364982605, "num_tokens": 55070620.0, "step": 1443 }, { "epoch": 0.18369164228469662, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.865494251251221, "learning_rate": 6.116998728274693e-07, "loss": 0.5062, "mean_token_accuracy": 0.8418616056442261, "num_tokens": 55108394.0, "step": 1444 }, { "epoch": 0.18381885256328712, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.904682159423828, "learning_rate": 6.121237812632472e-07, "loss": 0.5097, "mean_token_accuracy": 0.8344700336456299, "num_tokens": 55148170.0, "step": 1445 }, { "epoch": 0.18394606284187762, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 6.80167293548584, "learning_rate": 6.125476896990249e-07, "loss": 0.4523, "mean_token_accuracy": 0.8551996946334839, "num_tokens": 55188102.0, "step": 1446 }, { "epoch": 0.18407327312046814, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 6.848477840423584, "learning_rate": 6.129715981348028e-07, "loss": 0.4903, "mean_token_accuracy": 0.8460856676101685, "num_tokens": 55225893.0, "step": 1447 }, { "epoch": 0.18420048339905865, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.868958473205566, "learning_rate": 6.133955065705807e-07, "loss": 0.5105, "mean_token_accuracy": 0.8334200978279114, "num_tokens": 55265321.0, "step": 1448 }, { "epoch": 0.18432769367764915, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.853405475616455, "learning_rate": 6.138194150063585e-07, "loss": 0.4711, "mean_token_accuracy": 0.8479233980178833, "num_tokens": 55304179.0, "step": 1449 }, { "epoch": 0.18445490395623967, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.82378625869751, "learning_rate": 6.142433234421365e-07, "loss": 0.5049, "mean_token_accuracy": 0.8356788158416748, "num_tokens": 55343266.0, "step": 1450 }, { "epoch": 0.18458211423483017, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 6.841568946838379, "learning_rate": 6.146672318779143e-07, "loss": 0.5081, "mean_token_accuracy": 0.8366196155548096, "num_tokens": 55381147.0, "step": 1451 }, { "epoch": 0.18470932451342068, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.886171340942383, "learning_rate": 6.150911403136923e-07, "loss": 0.5084, "mean_token_accuracy": 0.8382409811019897, "num_tokens": 55414855.0, "step": 1452 }, { "epoch": 0.1848365347920112, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.825771808624268, "learning_rate": 6.155150487494701e-07, "loss": 0.4327, "mean_token_accuracy": 0.8592472076416016, "num_tokens": 55449699.0, "step": 1453 }, { "epoch": 0.1849637450706017, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 6.830165863037109, "learning_rate": 6.159389571852479e-07, "loss": 0.4601, "mean_token_accuracy": 0.8524476289749146, "num_tokens": 55493202.0, "step": 1454 }, { "epoch": 0.1850909553491922, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.806238651275635, "learning_rate": 6.163628656210258e-07, "loss": 0.4811, "mean_token_accuracy": 0.8471692204475403, "num_tokens": 55539873.0, "step": 1455 }, { "epoch": 0.18521816562778273, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.8661885261535645, "learning_rate": 6.167867740568037e-07, "loss": 0.5411, "mean_token_accuracy": 0.8284677267074585, "num_tokens": 55576338.0, "step": 1456 }, { "epoch": 0.18534537590637323, "ewc_loss": 0.0068359375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 6.845204830169678, "learning_rate": 6.172106824925815e-07, "loss": 0.462, "mean_token_accuracy": 0.8484236001968384, "num_tokens": 55618330.0, "step": 1457 }, { "epoch": 0.18547258618496373, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 6.83199405670166, "learning_rate": 6.176345909283595e-07, "loss": 0.4792, "mean_token_accuracy": 0.8445792198181152, "num_tokens": 55657394.0, "step": 1458 }, { "epoch": 0.18559979646355426, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 6.895453929901123, "learning_rate": 6.180584993641373e-07, "loss": 0.4646, "mean_token_accuracy": 0.8491030335426331, "num_tokens": 55695087.0, "step": 1459 }, { "epoch": 0.18572700674214476, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 6.8718743324279785, "learning_rate": 6.184824077999153e-07, "loss": 0.4493, "mean_token_accuracy": 0.8531942963600159, "num_tokens": 55730852.0, "step": 1460 }, { "epoch": 0.18585421702073526, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 6.931447505950928, "learning_rate": 6.189063162356931e-07, "loss": 0.5454, "mean_token_accuracy": 0.8266993165016174, "num_tokens": 55765678.0, "step": 1461 }, { "epoch": 0.1859814272993258, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.900588035583496, "learning_rate": 6.193302246714709e-07, "loss": 0.4713, "mean_token_accuracy": 0.8504400253295898, "num_tokens": 55804409.0, "step": 1462 }, { "epoch": 0.1861086375779163, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 6.883901119232178, "learning_rate": 6.197541331072488e-07, "loss": 0.5006, "mean_token_accuracy": 0.8376555442810059, "num_tokens": 55839520.0, "step": 1463 }, { "epoch": 0.1862358478565068, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 6.889158725738525, "learning_rate": 6.201780415430267e-07, "loss": 0.4806, "mean_token_accuracy": 0.8446731567382812, "num_tokens": 55879089.0, "step": 1464 }, { "epoch": 0.18636305813509732, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.926398754119873, "learning_rate": 6.206019499788045e-07, "loss": 0.523, "mean_token_accuracy": 0.8316266536712646, "num_tokens": 55916655.0, "step": 1465 }, { "epoch": 0.18649026841368782, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.870665550231934, "learning_rate": 6.210258584145825e-07, "loss": 0.4588, "mean_token_accuracy": 0.8479028940200806, "num_tokens": 55953828.0, "step": 1466 }, { "epoch": 0.18661747869227835, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 6.897712230682373, "learning_rate": 6.214497668503603e-07, "loss": 0.5076, "mean_token_accuracy": 0.8325533866882324, "num_tokens": 55992037.0, "step": 1467 }, { "epoch": 0.18674468897086885, "ewc_loss": 0.006988525390625, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.9654645919799805, "learning_rate": 6.218736752861383e-07, "loss": 0.468, "mean_token_accuracy": 0.8499524593353271, "num_tokens": 56020424.0, "step": 1468 }, { "epoch": 0.18687189924945935, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.921627521514893, "learning_rate": 6.22297583721916e-07, "loss": 0.5008, "mean_token_accuracy": 0.8436496257781982, "num_tokens": 56060839.0, "step": 1469 }, { "epoch": 0.18699910952804988, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.924544334411621, "learning_rate": 6.227214921576938e-07, "loss": 0.5309, "mean_token_accuracy": 0.8341203927993774, "num_tokens": 56099095.0, "step": 1470 }, { "epoch": 0.18712631980664038, "ewc_loss": 0.006988525390625, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.903922080993652, "learning_rate": 6.231454005934718e-07, "loss": 0.4269, "mean_token_accuracy": 0.8610994815826416, "num_tokens": 56133805.0, "step": 1471 }, { "epoch": 0.18725353008523088, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 6.95424747467041, "learning_rate": 6.235693090292496e-07, "loss": 0.5459, "mean_token_accuracy": 0.8288491368293762, "num_tokens": 56173064.0, "step": 1472 }, { "epoch": 0.1873807403638214, "ewc_loss": 0.006988525390625, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.8661789894104, "learning_rate": 6.239932174650275e-07, "loss": 0.4278, "mean_token_accuracy": 0.8626912832260132, "num_tokens": 56213285.0, "step": 1473 }, { "epoch": 0.1875079506424119, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 6.990426063537598, "learning_rate": 6.244171259008054e-07, "loss": 0.51, "mean_token_accuracy": 0.833736777305603, "num_tokens": 56248218.0, "step": 1474 }, { "epoch": 0.1876351609210024, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.866391658782959, "learning_rate": 6.248410343365833e-07, "loss": 0.4542, "mean_token_accuracy": 0.8530586361885071, "num_tokens": 56286098.0, "step": 1475 }, { "epoch": 0.18776237119959294, "ewc_loss": 0.006988525390625, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.881873607635498, "learning_rate": 6.252649427723612e-07, "loss": 0.4622, "mean_token_accuracy": 0.8523491621017456, "num_tokens": 56329732.0, "step": 1476 }, { "epoch": 0.18788958147818344, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.911468982696533, "learning_rate": 6.25688851208139e-07, "loss": 0.4671, "mean_token_accuracy": 0.8507349491119385, "num_tokens": 56367838.0, "step": 1477 }, { "epoch": 0.18801679175677394, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 6.901432514190674, "learning_rate": 6.261127596439168e-07, "loss": 0.5199, "mean_token_accuracy": 0.8320218920707703, "num_tokens": 56413059.0, "step": 1478 }, { "epoch": 0.18814400203536447, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.884099006652832, "learning_rate": 6.265366680796948e-07, "loss": 0.4224, "mean_token_accuracy": 0.857733964920044, "num_tokens": 56453806.0, "step": 1479 }, { "epoch": 0.18827121231395497, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.897911071777344, "learning_rate": 6.269605765154726e-07, "loss": 0.4704, "mean_token_accuracy": 0.845881462097168, "num_tokens": 56497448.0, "step": 1480 }, { "epoch": 0.18839842259254547, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.95170783996582, "learning_rate": 6.273844849512505e-07, "loss": 0.4726, "mean_token_accuracy": 0.8452531695365906, "num_tokens": 56537731.0, "step": 1481 }, { "epoch": 0.188525632871136, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.898621559143066, "learning_rate": 6.278083933870284e-07, "loss": 0.4117, "mean_token_accuracy": 0.8665652275085449, "num_tokens": 56580944.0, "step": 1482 }, { "epoch": 0.1886528431497265, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.9935526847839355, "learning_rate": 6.282323018228063e-07, "loss": 0.4443, "mean_token_accuracy": 0.8570741415023804, "num_tokens": 56616306.0, "step": 1483 }, { "epoch": 0.188780053428317, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 6.932079792022705, "learning_rate": 6.286562102585841e-07, "loss": 0.524, "mean_token_accuracy": 0.8351176977157593, "num_tokens": 56661874.0, "step": 1484 }, { "epoch": 0.18890726370690752, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.93175745010376, "learning_rate": 6.29080118694362e-07, "loss": 0.4189, "mean_token_accuracy": 0.8624163269996643, "num_tokens": 56696697.0, "step": 1485 }, { "epoch": 0.18903447398549802, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.931746959686279, "learning_rate": 6.295040271301398e-07, "loss": 0.4184, "mean_token_accuracy": 0.8609603047370911, "num_tokens": 56732581.0, "step": 1486 }, { "epoch": 0.18916168426408853, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 6.922417640686035, "learning_rate": 6.299279355659178e-07, "loss": 0.4677, "mean_token_accuracy": 0.8496014475822449, "num_tokens": 56777905.0, "step": 1487 }, { "epoch": 0.18928889454267905, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 6.9373860359191895, "learning_rate": 6.303518440016956e-07, "loss": 0.4869, "mean_token_accuracy": 0.84214848279953, "num_tokens": 56812915.0, "step": 1488 }, { "epoch": 0.18941610482126955, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 7.005377292633057, "learning_rate": 6.307757524374735e-07, "loss": 0.4819, "mean_token_accuracy": 0.8472078442573547, "num_tokens": 56846907.0, "step": 1489 }, { "epoch": 0.18954331509986008, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 6.9488115310668945, "learning_rate": 6.311996608732514e-07, "loss": 0.5104, "mean_token_accuracy": 0.8400498628616333, "num_tokens": 56884614.0, "step": 1490 }, { "epoch": 0.18967052537845058, "ewc_loss": 0.007049560546875, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 6.971423149108887, "learning_rate": 6.316235693090292e-07, "loss": 0.5013, "mean_token_accuracy": 0.8367744088172913, "num_tokens": 56920508.0, "step": 1491 }, { "epoch": 0.18979773565704108, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.962106227874756, "learning_rate": 6.320474777448071e-07, "loss": 0.5049, "mean_token_accuracy": 0.836428701877594, "num_tokens": 56957629.0, "step": 1492 }, { "epoch": 0.1899249459356316, "ewc_loss": 0.007080078125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 6.8869829177856445, "learning_rate": 6.324713861805849e-07, "loss": 0.4626, "mean_token_accuracy": 0.8476638793945312, "num_tokens": 56997265.0, "step": 1493 }, { "epoch": 0.1900521562142221, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 6.999152183532715, "learning_rate": 6.328952946163628e-07, "loss": 0.5651, "mean_token_accuracy": 0.8225612044334412, "num_tokens": 57031519.0, "step": 1494 }, { "epoch": 0.1901793664928126, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 6.961526393890381, "learning_rate": 6.333192030521407e-07, "loss": 0.436, "mean_token_accuracy": 0.8536480665206909, "num_tokens": 57067264.0, "step": 1495 }, { "epoch": 0.19030657677140314, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 6.91755485534668, "learning_rate": 6.337431114879186e-07, "loss": 0.4902, "mean_token_accuracy": 0.8428338170051575, "num_tokens": 57104906.0, "step": 1496 }, { "epoch": 0.19043378704999364, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 6.998533248901367, "learning_rate": 6.341670199236965e-07, "loss": 0.4571, "mean_token_accuracy": 0.8519700765609741, "num_tokens": 57142534.0, "step": 1497 }, { "epoch": 0.19056099732858414, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 6.907519340515137, "learning_rate": 6.345909283594744e-07, "loss": 0.4382, "mean_token_accuracy": 0.857085108757019, "num_tokens": 57188735.0, "step": 1498 }, { "epoch": 0.19068820760717467, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 6.979152679443359, "learning_rate": 6.350148367952522e-07, "loss": 0.4902, "mean_token_accuracy": 0.8355217576026917, "num_tokens": 57224610.0, "step": 1499 }, { "epoch": 0.19081541788576517, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 6.96226167678833, "learning_rate": 6.354387452310301e-07, "loss": 0.4454, "mean_token_accuracy": 0.8515019416809082, "num_tokens": 57259553.0, "step": 1500 }, { "epoch": 0.19094262816435567, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 6.9938740730285645, "learning_rate": 6.358626536668079e-07, "loss": 0.4492, "mean_token_accuracy": 0.8546188473701477, "num_tokens": 57295819.0, "step": 1501 }, { "epoch": 0.1910698384429462, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.3709068298339844e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 6.920028209686279, "learning_rate": 6.362865621025858e-07, "loss": 0.4377, "mean_token_accuracy": 0.8585658669471741, "num_tokens": 57339110.0, "step": 1502 }, { "epoch": 0.1911970487215367, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 7.016074180603027, "learning_rate": 6.367104705383637e-07, "loss": 0.4665, "mean_token_accuracy": 0.8485720753669739, "num_tokens": 57377311.0, "step": 1503 }, { "epoch": 0.1913242590001272, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 6.979049205780029, "learning_rate": 6.371343789741416e-07, "loss": 0.4789, "mean_token_accuracy": 0.84128737449646, "num_tokens": 57419281.0, "step": 1504 }, { "epoch": 0.19145146927871773, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 6.978094577789307, "learning_rate": 6.375582874099195e-07, "loss": 0.4666, "mean_token_accuracy": 0.8486478328704834, "num_tokens": 57458134.0, "step": 1505 }, { "epoch": 0.19157867955730823, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 7.0585784912109375, "learning_rate": 6.379821958456974e-07, "loss": 0.5207, "mean_token_accuracy": 0.8321609497070312, "num_tokens": 57492931.0, "step": 1506 }, { "epoch": 0.19170588983589873, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 6.963177680969238, "learning_rate": 6.384061042814751e-07, "loss": 0.4915, "mean_token_accuracy": 0.8397493362426758, "num_tokens": 57529581.0, "step": 1507 }, { "epoch": 0.19183310011448926, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 6.9907636642456055, "learning_rate": 6.38830012717253e-07, "loss": 0.4796, "mean_token_accuracy": 0.845032274723053, "num_tokens": 57570302.0, "step": 1508 }, { "epoch": 0.19196031039307976, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 6.948434829711914, "learning_rate": 6.392539211530309e-07, "loss": 0.4905, "mean_token_accuracy": 0.8393365144729614, "num_tokens": 57615974.0, "step": 1509 }, { "epoch": 0.19208752067167026, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.781650543212891e-06, "grad_norm": 6.978683948516846, "learning_rate": 6.396778295888087e-07, "loss": 0.4554, "mean_token_accuracy": 0.8543089032173157, "num_tokens": 57653737.0, "step": 1510 }, { "epoch": 0.1922147309502608, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 7.055245876312256, "learning_rate": 6.401017380245867e-07, "loss": 0.4891, "mean_token_accuracy": 0.8401476740837097, "num_tokens": 57693431.0, "step": 1511 }, { "epoch": 0.1923419412288513, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 6.9230146408081055, "learning_rate": 6.405256464603645e-07, "loss": 0.4735, "mean_token_accuracy": 0.844933032989502, "num_tokens": 57736413.0, "step": 1512 }, { "epoch": 0.1924691515074418, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 6.986149787902832, "learning_rate": 6.409495548961425e-07, "loss": 0.4579, "mean_token_accuracy": 0.8531184196472168, "num_tokens": 57773293.0, "step": 1513 }, { "epoch": 0.19259636178603232, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.0511651039123535, "learning_rate": 6.413734633319203e-07, "loss": 0.4702, "mean_token_accuracy": 0.8479832410812378, "num_tokens": 57809294.0, "step": 1514 }, { "epoch": 0.19272357206462282, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 6.9937238693237305, "learning_rate": 6.417973717676981e-07, "loss": 0.4416, "mean_token_accuracy": 0.8547543287277222, "num_tokens": 57843610.0, "step": 1515 }, { "epoch": 0.19285078234321335, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 7.007187843322754, "learning_rate": 6.42221280203476e-07, "loss": 0.5026, "mean_token_accuracy": 0.8344550728797913, "num_tokens": 57893098.0, "step": 1516 }, { "epoch": 0.19297799262180385, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 6.997704029083252, "learning_rate": 6.426451886392539e-07, "loss": 0.5284, "mean_token_accuracy": 0.8338432312011719, "num_tokens": 57933098.0, "step": 1517 }, { "epoch": 0.19310520290039435, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 7.0226640701293945, "learning_rate": 6.430690970750317e-07, "loss": 0.4597, "mean_token_accuracy": 0.8504282236099243, "num_tokens": 57971502.0, "step": 1518 }, { "epoch": 0.19323241317898487, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.0939249992370605, "learning_rate": 6.434930055108097e-07, "loss": 0.5092, "mean_token_accuracy": 0.8353614807128906, "num_tokens": 58001557.0, "step": 1519 }, { "epoch": 0.19335962345757537, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 7.0080342292785645, "learning_rate": 6.439169139465875e-07, "loss": 0.5183, "mean_token_accuracy": 0.8326594829559326, "num_tokens": 58038058.0, "step": 1520 }, { "epoch": 0.19348683373616588, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 6.998539924621582, "learning_rate": 6.443408223823655e-07, "loss": 0.4864, "mean_token_accuracy": 0.8413664102554321, "num_tokens": 58073138.0, "step": 1521 }, { "epoch": 0.1936140440147564, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.015758991241455, "learning_rate": 6.447647308181432e-07, "loss": 0.4779, "mean_token_accuracy": 0.844414234161377, "num_tokens": 58111793.0, "step": 1522 }, { "epoch": 0.1937412542933469, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 6.9671149253845215, "learning_rate": 6.451886392539211e-07, "loss": 0.4658, "mean_token_accuracy": 0.8477081060409546, "num_tokens": 58152911.0, "step": 1523 }, { "epoch": 0.1938684645719374, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 7.013283729553223, "learning_rate": 6.45612547689699e-07, "loss": 0.4834, "mean_token_accuracy": 0.8444130420684814, "num_tokens": 58192095.0, "step": 1524 }, { "epoch": 0.19399567485052793, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.027175426483154, "learning_rate": 6.460364561254769e-07, "loss": 0.4389, "mean_token_accuracy": 0.8556669354438782, "num_tokens": 58232651.0, "step": 1525 }, { "epoch": 0.19412288512911843, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 7.000851631164551, "learning_rate": 6.464603645612547e-07, "loss": 0.4365, "mean_token_accuracy": 0.8584827184677124, "num_tokens": 58270802.0, "step": 1526 }, { "epoch": 0.19425009540770893, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 6.991827011108398, "learning_rate": 6.468842729970327e-07, "loss": 0.4682, "mean_token_accuracy": 0.8505716919898987, "num_tokens": 58313179.0, "step": 1527 }, { "epoch": 0.19437730568629946, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 7.033352375030518, "learning_rate": 6.473081814328105e-07, "loss": 0.5081, "mean_token_accuracy": 0.8407294750213623, "num_tokens": 58354728.0, "step": 1528 }, { "epoch": 0.19450451596488996, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.000431060791016, "learning_rate": 6.477320898685885e-07, "loss": 0.4913, "mean_token_accuracy": 0.8439478874206543, "num_tokens": 58398649.0, "step": 1529 }, { "epoch": 0.19463172624348046, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.020411491394043, "learning_rate": 6.481559983043662e-07, "loss": 0.5027, "mean_token_accuracy": 0.8370258808135986, "num_tokens": 58438261.0, "step": 1530 }, { "epoch": 0.194758936522071, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 7.023792266845703, "learning_rate": 6.48579906740144e-07, "loss": 0.4475, "mean_token_accuracy": 0.8532919883728027, "num_tokens": 58479141.0, "step": 1531 }, { "epoch": 0.1948861468006615, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.018093585968018, "learning_rate": 6.49003815175922e-07, "loss": 0.4158, "mean_token_accuracy": 0.8648521304130554, "num_tokens": 58515700.0, "step": 1532 }, { "epoch": 0.195013357079252, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.081475257873535, "learning_rate": 6.494277236116998e-07, "loss": 0.4224, "mean_token_accuracy": 0.8630925416946411, "num_tokens": 58549825.0, "step": 1533 }, { "epoch": 0.19514056735784252, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.020635604858398, "learning_rate": 6.498516320474777e-07, "loss": 0.5621, "mean_token_accuracy": 0.8236045241355896, "num_tokens": 58596110.0, "step": 1534 }, { "epoch": 0.19526777763643302, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.072238922119141, "learning_rate": 6.502755404832556e-07, "loss": 0.4724, "mean_token_accuracy": 0.8448919653892517, "num_tokens": 58632410.0, "step": 1535 }, { "epoch": 0.19539498791502352, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.07804536819458, "learning_rate": 6.506994489190335e-07, "loss": 0.534, "mean_token_accuracy": 0.8333945274353027, "num_tokens": 58671339.0, "step": 1536 }, { "epoch": 0.19552219819361405, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.053544521331787, "learning_rate": 6.511233573548114e-07, "loss": 0.4341, "mean_token_accuracy": 0.8571967482566833, "num_tokens": 58711786.0, "step": 1537 }, { "epoch": 0.19564940847220455, "ewc_loss": 0.007354736328125, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 7.1125335693359375, "learning_rate": 6.515472657905892e-07, "loss": 0.4775, "mean_token_accuracy": 0.8454220294952393, "num_tokens": 58743600.0, "step": 1538 }, { "epoch": 0.19577661875079505, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.134649276733398, "learning_rate": 6.51971174226367e-07, "loss": 0.5336, "mean_token_accuracy": 0.8290468454360962, "num_tokens": 58780622.0, "step": 1539 }, { "epoch": 0.19590382902938558, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.040106773376465, "learning_rate": 6.52395082662145e-07, "loss": 0.5265, "mean_token_accuracy": 0.8307855725288391, "num_tokens": 58819370.0, "step": 1540 }, { "epoch": 0.19603103930797608, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 7.084467887878418, "learning_rate": 6.528189910979228e-07, "loss": 0.451, "mean_token_accuracy": 0.8533893823623657, "num_tokens": 58856270.0, "step": 1541 }, { "epoch": 0.1961582495865666, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.014833927154541, "learning_rate": 6.532428995337007e-07, "loss": 0.4681, "mean_token_accuracy": 0.8478410243988037, "num_tokens": 58898702.0, "step": 1542 }, { "epoch": 0.1962854598651571, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.061440467834473, "learning_rate": 6.536668079694786e-07, "loss": 0.4895, "mean_token_accuracy": 0.8392990231513977, "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": 7.091311454772949, "learning_rate": 6.540907164052565e-07, "loss": 0.4759, "mean_token_accuracy": 0.8445411920547485, "num_tokens": 58977766.0, "step": 1544 }, { "epoch": 0.19653988042233814, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.038145065307617, "learning_rate": 6.545146248410343e-07, "loss": 0.4855, "mean_token_accuracy": 0.8426161408424377, "num_tokens": 59019649.0, "step": 1545 }, { "epoch": 0.19666709070092864, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.11851167678833, "learning_rate": 6.549385332768122e-07, "loss": 0.4977, "mean_token_accuracy": 0.8412085175514221, "num_tokens": 59060543.0, "step": 1546 }, { "epoch": 0.19679430097951914, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.124604225158691, "learning_rate": 6.5536244171259e-07, "loss": 0.4965, "mean_token_accuracy": 0.8383398056030273, "num_tokens": 59100049.0, "step": 1547 }, { "epoch": 0.19692151125810967, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.097812175750732, "learning_rate": 6.55786350148368e-07, "loss": 0.5027, "mean_token_accuracy": 0.837239146232605, "num_tokens": 59137370.0, "step": 1548 }, { "epoch": 0.19704872153670017, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.1200785636901855, "learning_rate": 6.562102585841458e-07, "loss": 0.4557, "mean_token_accuracy": 0.8467439413070679, "num_tokens": 59176196.0, "step": 1549 }, { "epoch": 0.19717593181529067, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.059238910675049, "learning_rate": 6.566341670199236e-07, "loss": 0.4543, "mean_token_accuracy": 0.8531985878944397, "num_tokens": 59216215.0, "step": 1550 }, { "epoch": 0.1973031420938812, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.020069122314453e-06, "grad_norm": 7.146495819091797, "learning_rate": 6.570580754557016e-07, "loss": 0.4703, "mean_token_accuracy": 0.8473958969116211, "num_tokens": 59254173.0, "step": 1551 }, { "epoch": 0.1974303523724717, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.13143253326416, "learning_rate": 6.574819838914794e-07, "loss": 0.4427, "mean_token_accuracy": 0.8555363416671753, "num_tokens": 59288496.0, "step": 1552 }, { "epoch": 0.1975575626510622, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.1597819328308105, "learning_rate": 6.579058923272573e-07, "loss": 0.5011, "mean_token_accuracy": 0.8371906280517578, "num_tokens": 59329043.0, "step": 1553 }, { "epoch": 0.19768477292965272, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.135304927825928, "learning_rate": 6.583298007630351e-07, "loss": 0.4919, "mean_token_accuracy": 0.8417929410934448, "num_tokens": 59360743.0, "step": 1554 }, { "epoch": 0.19781198320824323, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.107850551605225, "learning_rate": 6.58753709198813e-07, "loss": 0.4886, "mean_token_accuracy": 0.8414241075515747, "num_tokens": 59404167.0, "step": 1555 }, { "epoch": 0.19793919348683373, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.1153764724731445, "learning_rate": 6.591776176345909e-07, "loss": 0.4564, "mean_token_accuracy": 0.8484756946563721, "num_tokens": 59447770.0, "step": 1556 }, { "epoch": 0.19806640376542425, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.120079517364502, "learning_rate": 6.596015260703688e-07, "loss": 0.4656, "mean_token_accuracy": 0.8524296283721924, "num_tokens": 59494989.0, "step": 1557 }, { "epoch": 0.19819361404401475, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.106777667999268, "learning_rate": 6.600254345061466e-07, "loss": 0.4531, "mean_token_accuracy": 0.854301393032074, "num_tokens": 59534926.0, "step": 1558 }, { "epoch": 0.19832082432260525, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.131320476531982, "learning_rate": 6.604493429419246e-07, "loss": 0.5079, "mean_token_accuracy": 0.836932897567749, "num_tokens": 59577832.0, "step": 1559 }, { "epoch": 0.19844803460119578, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.110631942749023, "learning_rate": 6.608732513777023e-07, "loss": 0.5153, "mean_token_accuracy": 0.8416455984115601, "num_tokens": 59624995.0, "step": 1560 }, { "epoch": 0.19857524487978628, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 7.119132041931152, "learning_rate": 6.612971598134803e-07, "loss": 0.4685, "mean_token_accuracy": 0.8470832109451294, "num_tokens": 59663000.0, "step": 1561 }, { "epoch": 0.19870245515837678, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.175594806671143, "learning_rate": 6.617210682492581e-07, "loss": 0.5225, "mean_token_accuracy": 0.8322625160217285, "num_tokens": 59700241.0, "step": 1562 }, { "epoch": 0.1988296654369673, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.129007816314697, "learning_rate": 6.62144976685036e-07, "loss": 0.4416, "mean_token_accuracy": 0.858415424823761, "num_tokens": 59737244.0, "step": 1563 }, { "epoch": 0.1989568757155578, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.147036075592041, "learning_rate": 6.625688851208139e-07, "loss": 0.4914, "mean_token_accuracy": 0.8423236608505249, "num_tokens": 59775538.0, "step": 1564 }, { "epoch": 0.19908408599414834, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.170684337615967, "learning_rate": 6.629927935565918e-07, "loss": 0.4863, "mean_token_accuracy": 0.8400726318359375, "num_tokens": 59810373.0, "step": 1565 }, { "epoch": 0.19921129627273884, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.157248497009277, "learning_rate": 6.634167019923696e-07, "loss": 0.5178, "mean_token_accuracy": 0.8379891514778137, "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": 7.146813869476318, "learning_rate": 6.638406104281476e-07, "loss": 0.4785, "mean_token_accuracy": 0.8444836139678955, "num_tokens": 59882552.0, "step": 1567 }, { "epoch": 0.19946571682991987, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 7.136019706726074, "learning_rate": 6.642645188639253e-07, "loss": 0.4539, "mean_token_accuracy": 0.8508783578872681, "num_tokens": 59925090.0, "step": 1568 }, { "epoch": 0.19959292710851037, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.178235054016113, "learning_rate": 6.646884272997032e-07, "loss": 0.5017, "mean_token_accuracy": 0.837898850440979, "num_tokens": 59965406.0, "step": 1569 }, { "epoch": 0.19972013738710087, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.175540447235107, "learning_rate": 6.651123357354811e-07, "loss": 0.4887, "mean_token_accuracy": 0.8423380851745605, "num_tokens": 60001770.0, "step": 1570 }, { "epoch": 0.1998473476656914, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.174018383026123, "learning_rate": 6.655362441712589e-07, "loss": 0.5054, "mean_token_accuracy": 0.835808515548706, "num_tokens": 60036911.0, "step": 1571 }, { "epoch": 0.1999745579442819, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.127123832702637, "learning_rate": 6.659601526070369e-07, "loss": 0.5327, "mean_token_accuracy": 0.8323330879211426, "num_tokens": 60080210.0, "step": 1572 }, { "epoch": 0.2001017682228724, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 7.230222702026367, "learning_rate": 6.663840610428147e-07, "loss": 0.567, "mean_token_accuracy": 0.82415771484375, "num_tokens": 60111533.0, "step": 1573 }, { "epoch": 0.20022897850146293, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.170193195343018, "learning_rate": 6.668079694785926e-07, "loss": 0.4443, "mean_token_accuracy": 0.8558433055877686, "num_tokens": 60150733.0, "step": 1574 }, { "epoch": 0.20035618878005343, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.079673767089844e-06, "grad_norm": 7.13472843170166, "learning_rate": 6.672318779143704e-07, "loss": 0.4637, "mean_token_accuracy": 0.850069522857666, "num_tokens": 60185034.0, "step": 1575 }, { "epoch": 0.20048339905864393, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.232385158538818, "learning_rate": 6.676557863501483e-07, "loss": 0.4947, "mean_token_accuracy": 0.8405188322067261, "num_tokens": 60216865.0, "step": 1576 }, { "epoch": 0.20061060933723446, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.128419399261475, "learning_rate": 6.680796947859262e-07, "loss": 0.4834, "mean_token_accuracy": 0.84287428855896, "num_tokens": 60257685.0, "step": 1577 }, { "epoch": 0.20073781961582496, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 7.180855751037598, "learning_rate": 6.685036032217041e-07, "loss": 0.4782, "mean_token_accuracy": 0.8455037474632263, "num_tokens": 60294258.0, "step": 1578 }, { "epoch": 0.20086502989441546, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.168710708618164, "learning_rate": 6.689275116574819e-07, "loss": 0.4681, "mean_token_accuracy": 0.8512681722640991, "num_tokens": 60330319.0, "step": 1579 }, { "epoch": 0.200992240173006, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 7.160706520080566, "learning_rate": 6.693514200932599e-07, "loss": 0.4849, "mean_token_accuracy": 0.8457220792770386, "num_tokens": 60370482.0, "step": 1580 }, { "epoch": 0.2011194504515965, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.154212474822998, "learning_rate": 6.697753285290377e-07, "loss": 0.4902, "mean_token_accuracy": 0.841015100479126, "num_tokens": 60412436.0, "step": 1581 }, { "epoch": 0.201246660730187, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.204110145568848, "learning_rate": 6.701992369648156e-07, "loss": 0.4139, "mean_token_accuracy": 0.8659083843231201, "num_tokens": 60448982.0, "step": 1582 }, { "epoch": 0.20137387100877752, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.1232733726501465, "learning_rate": 6.706231454005934e-07, "loss": 0.431, "mean_token_accuracy": 0.8573634624481201, "num_tokens": 60489878.0, "step": 1583 }, { "epoch": 0.20150108128736802, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.2083740234375, "learning_rate": 6.710470538363713e-07, "loss": 0.4662, "mean_token_accuracy": 0.8508812189102173, "num_tokens": 60528979.0, "step": 1584 }, { "epoch": 0.20162829156595852, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.204211235046387, "learning_rate": 6.714709622721492e-07, "loss": 0.495, "mean_token_accuracy": 0.8394992351531982, "num_tokens": 60561670.0, "step": 1585 }, { "epoch": 0.20175550184454905, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.161705493927002, "learning_rate": 6.718948707079271e-07, "loss": 0.4693, "mean_token_accuracy": 0.8455570936203003, "num_tokens": 60602061.0, "step": 1586 }, { "epoch": 0.20188271212313955, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.203519344329834, "learning_rate": 6.723187791437049e-07, "loss": 0.472, "mean_token_accuracy": 0.8470480442047119, "num_tokens": 60639805.0, "step": 1587 }, { "epoch": 0.20200992240173005, "ewc_loss": 0.007720947265625, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.191850185394287, "learning_rate": 6.727426875794829e-07, "loss": 0.4988, "mean_token_accuracy": 0.8401094675064087, "num_tokens": 60681816.0, "step": 1588 }, { "epoch": 0.20213713268032057, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.185193061828613, "learning_rate": 6.731665960152607e-07, "loss": 0.4832, "mean_token_accuracy": 0.8405845761299133, "num_tokens": 60726343.0, "step": 1589 }, { "epoch": 0.20226434295891108, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.237691879272461, "learning_rate": 6.735905044510385e-07, "loss": 0.5276, "mean_token_accuracy": 0.82468181848526, "num_tokens": 60764814.0, "step": 1590 }, { "epoch": 0.2023915532375016, "ewc_loss": 0.007720947265625, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.21366548538208, "learning_rate": 6.740144128868164e-07, "loss": 0.4947, "mean_token_accuracy": 0.841576337814331, "num_tokens": 60802384.0, "step": 1591 }, { "epoch": 0.2025187635160921, "ewc_loss": 0.007720947265625, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 7.206951141357422, "learning_rate": 6.744383213225942e-07, "loss": 0.5206, "mean_token_accuracy": 0.8313615322113037, "num_tokens": 60842085.0, "step": 1592 }, { "epoch": 0.2026459737946826, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.23383903503418, "learning_rate": 6.748622297583722e-07, "loss": 0.4565, "mean_token_accuracy": 0.8469923138618469, "num_tokens": 60874835.0, "step": 1593 }, { "epoch": 0.20277318407327313, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.1755170822143555, "learning_rate": 6.7528613819415e-07, "loss": 0.4823, "mean_token_accuracy": 0.8415091633796692, "num_tokens": 60910682.0, "step": 1594 }, { "epoch": 0.20290039435186363, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 7.187715530395508, "learning_rate": 6.757100466299279e-07, "loss": 0.479, "mean_token_accuracy": 0.8467882871627808, "num_tokens": 60951352.0, "step": 1595 }, { "epoch": 0.20302760463045413, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 7.316068172454834, "learning_rate": 6.761339550657058e-07, "loss": 0.5067, "mean_token_accuracy": 0.8335844874382019, "num_tokens": 60979055.0, "step": 1596 }, { "epoch": 0.20315481490904466, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.4722347259521484e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.156391143798828, "learning_rate": 6.765578635014837e-07, "loss": 0.486, "mean_token_accuracy": 0.8454197645187378, "num_tokens": 61017437.0, "step": 1597 }, { "epoch": 0.20328202518763516, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.4722347259521484e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 7.202714443206787, "learning_rate": 6.769817719372614e-07, "loss": 0.4616, "mean_token_accuracy": 0.8504636287689209, "num_tokens": 61052286.0, "step": 1598 }, { "epoch": 0.20340923546622566, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4722347259521484e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.249049186706543, "learning_rate": 6.774056803730394e-07, "loss": 0.5071, "mean_token_accuracy": 0.8361607193946838, "num_tokens": 61086110.0, "step": 1599 }, { "epoch": 0.2035364457448162, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.183104991912842, "learning_rate": 6.778295888088172e-07, "loss": 0.4783, "mean_token_accuracy": 0.8449598550796509, "num_tokens": 61124181.0, "step": 1600 }, { "epoch": 0.2036636560234067, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 7.236949443817139, "learning_rate": 6.782534972445952e-07, "loss": 0.4986, "mean_token_accuracy": 0.8376085758209229, "num_tokens": 61160881.0, "step": 1601 }, { "epoch": 0.2037908663019972, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.286804676055908, "learning_rate": 6.78677405680373e-07, "loss": 0.4966, "mean_token_accuracy": 0.8414263129234314, "num_tokens": 61200755.0, "step": 1602 }, { "epoch": 0.20391807658058772, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.206082820892334, "learning_rate": 6.791013141161509e-07, "loss": 0.487, "mean_token_accuracy": 0.8407508134841919, "num_tokens": 61240045.0, "step": 1603 }, { "epoch": 0.20404528685917822, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.207983493804932, "learning_rate": 6.795252225519288e-07, "loss": 0.5124, "mean_token_accuracy": 0.832569420337677, "num_tokens": 61282832.0, "step": 1604 }, { "epoch": 0.20417249713776872, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.237510681152344, "learning_rate": 6.799491309877067e-07, "loss": 0.4862, "mean_token_accuracy": 0.8400866985321045, "num_tokens": 61319841.0, "step": 1605 }, { "epoch": 0.20429970741635925, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.273297309875488, "learning_rate": 6.803730394234844e-07, "loss": 0.4547, "mean_token_accuracy": 0.8526641130447388, "num_tokens": 61352732.0, "step": 1606 }, { "epoch": 0.20442691769494975, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.259057998657227, "learning_rate": 6.807969478592624e-07, "loss": 0.4882, "mean_token_accuracy": 0.8440030813217163, "num_tokens": 61388324.0, "step": 1607 }, { "epoch": 0.20455412797354025, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 7.242406368255615, "learning_rate": 6.812208562950402e-07, "loss": 0.5118, "mean_token_accuracy": 0.8409372568130493, "num_tokens": 61431222.0, "step": 1608 }, { "epoch": 0.20468133825213078, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.2075066566467285, "learning_rate": 6.816447647308182e-07, "loss": 0.5418, "mean_token_accuracy": 0.8266981840133667, "num_tokens": 61475617.0, "step": 1609 }, { "epoch": 0.20480854853072128, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.2660088539123535, "learning_rate": 6.82068673166596e-07, "loss": 0.4683, "mean_token_accuracy": 0.8491834402084351, "num_tokens": 61511338.0, "step": 1610 }, { "epoch": 0.20493575880931178, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.221984386444092, "learning_rate": 6.824925816023738e-07, "loss": 0.4904, "mean_token_accuracy": 0.839870810508728, "num_tokens": 61549093.0, "step": 1611 }, { "epoch": 0.2050629690879023, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.194299221038818, "learning_rate": 6.829164900381518e-07, "loss": 0.4584, "mean_token_accuracy": 0.8504283428192139, "num_tokens": 61593843.0, "step": 1612 }, { "epoch": 0.2051901793664928, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 7.224647045135498, "learning_rate": 6.833403984739295e-07, "loss": 0.4694, "mean_token_accuracy": 0.8471446633338928, "num_tokens": 61629362.0, "step": 1613 }, { "epoch": 0.2053173896450833, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.263174057006836, "learning_rate": 6.837643069097074e-07, "loss": 0.5505, "mean_token_accuracy": 0.824405312538147, "num_tokens": 61667037.0, "step": 1614 }, { "epoch": 0.20544459992367384, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.254185676574707, "learning_rate": 6.841882153454853e-07, "loss": 0.4675, "mean_token_accuracy": 0.8459159135818481, "num_tokens": 61701976.0, "step": 1615 }, { "epoch": 0.20557181020226434, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.250570297241211, "learning_rate": 6.846121237812632e-07, "loss": 0.4363, "mean_token_accuracy": 0.8577607870101929, "num_tokens": 61739068.0, "step": 1616 }, { "epoch": 0.20569902048085487, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.222601890563965, "learning_rate": 6.850360322170411e-07, "loss": 0.4529, "mean_token_accuracy": 0.8531254529953003, "num_tokens": 61781515.0, "step": 1617 }, { "epoch": 0.20582623075944537, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.274109363555908, "learning_rate": 6.85459940652819e-07, "loss": 0.4665, "mean_token_accuracy": 0.8461745381355286, "num_tokens": 61815201.0, "step": 1618 }, { "epoch": 0.20595344103803587, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.253168106079102, "learning_rate": 6.858838490885968e-07, "loss": 0.5226, "mean_token_accuracy": 0.8316027522087097, "num_tokens": 61856639.0, "step": 1619 }, { "epoch": 0.2060806513166264, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.285652160644531, "learning_rate": 6.863077575243748e-07, "loss": 0.535, "mean_token_accuracy": 0.8303959369659424, "num_tokens": 61894313.0, "step": 1620 }, { "epoch": 0.2062078615952169, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.339186668395996, "learning_rate": 6.867316659601525e-07, "loss": 0.4706, "mean_token_accuracy": 0.8453655242919922, "num_tokens": 61930748.0, "step": 1621 }, { "epoch": 0.2063350718738074, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.262932777404785, "learning_rate": 6.871555743959304e-07, "loss": 0.4855, "mean_token_accuracy": 0.8443576097488403, "num_tokens": 61972002.0, "step": 1622 }, { "epoch": 0.20646228215239792, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.278070449829102, "learning_rate": 6.875794828317083e-07, "loss": 0.4617, "mean_token_accuracy": 0.8511592745780945, "num_tokens": 62010141.0, "step": 1623 }, { "epoch": 0.20658949243098843, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.262823581695557, "learning_rate": 6.880033912674862e-07, "loss": 0.4496, "mean_token_accuracy": 0.854804515838623, "num_tokens": 62055907.0, "step": 1624 }, { "epoch": 0.20671670270957893, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.310466766357422, "learning_rate": 6.884272997032641e-07, "loss": 0.5563, "mean_token_accuracy": 0.8221930265426636, "num_tokens": 62095885.0, "step": 1625 }, { "epoch": 0.20684391298816945, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 7.3001203536987305, "learning_rate": 6.88851208139042e-07, "loss": 0.4491, "mean_token_accuracy": 0.8559812307357788, "num_tokens": 62130497.0, "step": 1626 }, { "epoch": 0.20697112326675995, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.292253017425537, "learning_rate": 6.892751165748198e-07, "loss": 0.5085, "mean_token_accuracy": 0.8339651226997375, "num_tokens": 62170296.0, "step": 1627 }, { "epoch": 0.20709833354535045, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 7.256910800933838, "learning_rate": 6.896990250105978e-07, "loss": 0.446, "mean_token_accuracy": 0.8587378263473511, "num_tokens": 62210168.0, "step": 1628 }, { "epoch": 0.20722554382394098, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.311554908752441, "learning_rate": 6.901229334463755e-07, "loss": 0.4714, "mean_token_accuracy": 0.8456428647041321, "num_tokens": 62245251.0, "step": 1629 }, { "epoch": 0.20735275410253148, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.285619258880615, "learning_rate": 6.905468418821534e-07, "loss": 0.4699, "mean_token_accuracy": 0.8519418835639954, "num_tokens": 62286124.0, "step": 1630 }, { "epoch": 0.20747996438112198, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.30021333694458, "learning_rate": 6.909707503179313e-07, "loss": 0.4746, "mean_token_accuracy": 0.8436176776885986, "num_tokens": 62325604.0, "step": 1631 }, { "epoch": 0.2076071746597125, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.2709126472473145, "learning_rate": 6.913946587537091e-07, "loss": 0.4704, "mean_token_accuracy": 0.8470960855484009, "num_tokens": 62364467.0, "step": 1632 }, { "epoch": 0.207734384938303, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.3360419273376465, "learning_rate": 6.918185671894871e-07, "loss": 0.544, "mean_token_accuracy": 0.8261491060256958, "num_tokens": 62402075.0, "step": 1633 }, { "epoch": 0.2078615952168935, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.339217185974121, "learning_rate": 6.922424756252649e-07, "loss": 0.5113, "mean_token_accuracy": 0.8419771194458008, "num_tokens": 62437802.0, "step": 1634 }, { "epoch": 0.20798880549548404, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.287260055541992, "learning_rate": 6.926663840610428e-07, "loss": 0.4297, "mean_token_accuracy": 0.8585982322692871, "num_tokens": 62476176.0, "step": 1635 }, { "epoch": 0.20811601577407454, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 7.290703296661377, "learning_rate": 6.930902924968206e-07, "loss": 0.4742, "mean_token_accuracy": 0.8424135446548462, "num_tokens": 62514378.0, "step": 1636 }, { "epoch": 0.20824322605266504, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.322714328765869, "learning_rate": 6.935142009325985e-07, "loss": 0.4962, "mean_token_accuracy": 0.838982343673706, "num_tokens": 62555302.0, "step": 1637 }, { "epoch": 0.20837043633125557, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.305248737335205, "learning_rate": 6.939381093683764e-07, "loss": 0.515, "mean_token_accuracy": 0.8341577649116516, "num_tokens": 62594356.0, "step": 1638 }, { "epoch": 0.20849764660984607, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.314034461975098, "learning_rate": 6.943620178041543e-07, "loss": 0.4328, "mean_token_accuracy": 0.8597777485847473, "num_tokens": 62632903.0, "step": 1639 }, { "epoch": 0.2086248568884366, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 7.316752910614014, "learning_rate": 6.947859262399321e-07, "loss": 0.4828, "mean_token_accuracy": 0.8499881029129028, "num_tokens": 62668979.0, "step": 1640 }, { "epoch": 0.2087520671670271, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.338203430175781, "learning_rate": 6.952098346757101e-07, "loss": 0.4687, "mean_token_accuracy": 0.8474273681640625, "num_tokens": 62705867.0, "step": 1641 }, { "epoch": 0.2088792774456176, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.338873386383057, "learning_rate": 6.956337431114879e-07, "loss": 0.4643, "mean_token_accuracy": 0.8505733013153076, "num_tokens": 62744559.0, "step": 1642 }, { "epoch": 0.20900648772420813, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.3159565925598145, "learning_rate": 6.960576515472658e-07, "loss": 0.4956, "mean_token_accuracy": 0.8411974906921387, "num_tokens": 62785221.0, "step": 1643 }, { "epoch": 0.20913369800279863, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.364743232727051, "learning_rate": 6.964815599830436e-07, "loss": 0.4853, "mean_token_accuracy": 0.8430438041687012, "num_tokens": 62820095.0, "step": 1644 }, { "epoch": 0.20926090828138913, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.312644004821777, "learning_rate": 6.969054684188215e-07, "loss": 0.5358, "mean_token_accuracy": 0.8240368962287903, "num_tokens": 62863414.0, "step": 1645 }, { "epoch": 0.20938811855997966, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.344438552856445, "learning_rate": 6.973293768545994e-07, "loss": 0.4458, "mean_token_accuracy": 0.8549321889877319, "num_tokens": 62904647.0, "step": 1646 }, { "epoch": 0.20951532883857016, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.294585704803467, "learning_rate": 6.977532852903773e-07, "loss": 0.4791, "mean_token_accuracy": 0.8431650996208191, "num_tokens": 62949476.0, "step": 1647 }, { "epoch": 0.20964253911716066, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.3446760177612305, "learning_rate": 6.981771937261551e-07, "loss": 0.5319, "mean_token_accuracy": 0.8300583362579346, "num_tokens": 62984784.0, "step": 1648 }, { "epoch": 0.2097697493957512, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.376476764678955, "learning_rate": 6.986011021619331e-07, "loss": 0.4863, "mean_token_accuracy": 0.8428831100463867, "num_tokens": 63024395.0, "step": 1649 }, { "epoch": 0.2098969596743417, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 7.387695789337158, "learning_rate": 6.990250105977109e-07, "loss": 0.5021, "mean_token_accuracy": 0.8399327397346497, "num_tokens": 63061886.0, "step": 1650 }, { "epoch": 0.2100241699529322, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.347228527069092, "learning_rate": 6.994489190334886e-07, "loss": 0.4703, "mean_token_accuracy": 0.8456768989562988, "num_tokens": 63104150.0, "step": 1651 }, { "epoch": 0.21015138023152272, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.298289775848389, "learning_rate": 6.998728274692666e-07, "loss": 0.4338, "mean_token_accuracy": 0.8580245971679688, "num_tokens": 63144346.0, "step": 1652 }, { "epoch": 0.21027859051011322, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 7.3915486335754395, "learning_rate": 7.002967359050444e-07, "loss": 0.476, "mean_token_accuracy": 0.8516627550125122, "num_tokens": 63180042.0, "step": 1653 }, { "epoch": 0.21040580078870372, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.421504974365234, "learning_rate": 7.007206443408224e-07, "loss": 0.4258, "mean_token_accuracy": 0.8592344522476196, "num_tokens": 63216570.0, "step": 1654 }, { "epoch": 0.21053301106729425, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.332622051239014, "learning_rate": 7.011445527766002e-07, "loss": 0.4775, "mean_token_accuracy": 0.8453017473220825, "num_tokens": 63253457.0, "step": 1655 }, { "epoch": 0.21066022134588475, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.334192276000977, "learning_rate": 7.015684612123781e-07, "loss": 0.4627, "mean_token_accuracy": 0.8483681678771973, "num_tokens": 63295430.0, "step": 1656 }, { "epoch": 0.21078743162447525, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.320506572723389, "learning_rate": 7.01992369648156e-07, "loss": 0.4847, "mean_token_accuracy": 0.845446765422821, "num_tokens": 63340118.0, "step": 1657 }, { "epoch": 0.21091464190306577, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.408280849456787, "learning_rate": 7.024162780839339e-07, "loss": 0.5159, "mean_token_accuracy": 0.8267467021942139, "num_tokens": 63372037.0, "step": 1658 }, { "epoch": 0.21104185218165628, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.343243598937988, "learning_rate": 7.028401865197116e-07, "loss": 0.5065, "mean_token_accuracy": 0.83553546667099, "num_tokens": 63415508.0, "step": 1659 }, { "epoch": 0.21116906246024678, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.274725914001465, "learning_rate": 7.032640949554896e-07, "loss": 0.4157, "mean_token_accuracy": 0.860049843788147, "num_tokens": 63459320.0, "step": 1660 }, { "epoch": 0.2112962727388373, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 7.385161876678467, "learning_rate": 7.036880033912674e-07, "loss": 0.5149, "mean_token_accuracy": 0.8351773023605347, "num_tokens": 63499624.0, "step": 1661 }, { "epoch": 0.2114234830174278, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.3651556968688965, "learning_rate": 7.041119118270454e-07, "loss": 0.4714, "mean_token_accuracy": 0.8459615707397461, "num_tokens": 63536507.0, "step": 1662 }, { "epoch": 0.2115506932960183, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 7.351531505584717, "learning_rate": 7.045358202628232e-07, "loss": 0.5125, "mean_token_accuracy": 0.8365147709846497, "num_tokens": 63576483.0, "step": 1663 }, { "epoch": 0.21167790357460883, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.360677719116211, "learning_rate": 7.049597286986011e-07, "loss": 0.5021, "mean_token_accuracy": 0.8366955518722534, "num_tokens": 63615568.0, "step": 1664 }, { "epoch": 0.21180511385319933, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.37764835357666, "learning_rate": 7.05383637134379e-07, "loss": 0.4628, "mean_token_accuracy": 0.8497627973556519, "num_tokens": 63650023.0, "step": 1665 }, { "epoch": 0.21193232413178986, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.370049953460693, "learning_rate": 7.058075455701568e-07, "loss": 0.4932, "mean_token_accuracy": 0.8360366821289062, "num_tokens": 63690229.0, "step": 1666 }, { "epoch": 0.21205953441038036, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.337626934051514, "learning_rate": 7.062314540059346e-07, "loss": 0.4376, "mean_token_accuracy": 0.8594331741333008, "num_tokens": 63728169.0, "step": 1667 }, { "epoch": 0.21218674468897086, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.4117112159729, "learning_rate": 7.066553624417126e-07, "loss": 0.4626, "mean_token_accuracy": 0.8477674722671509, "num_tokens": 63762816.0, "step": 1668 }, { "epoch": 0.2123139549675614, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.40986967086792, "learning_rate": 7.070792708774904e-07, "loss": 0.4841, "mean_token_accuracy": 0.8436827659606934, "num_tokens": 63804288.0, "step": 1669 }, { "epoch": 0.2124411652461519, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.325090408325195, "learning_rate": 7.075031793132684e-07, "loss": 0.5052, "mean_token_accuracy": 0.8411937952041626, "num_tokens": 63851365.0, "step": 1670 }, { "epoch": 0.2125683755247424, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 7.418638229370117, "learning_rate": 7.079270877490462e-07, "loss": 0.4666, "mean_token_accuracy": 0.8499504923820496, "num_tokens": 63886187.0, "step": 1671 }, { "epoch": 0.21269558580333292, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.412410736083984, "learning_rate": 7.08350996184824e-07, "loss": 0.5402, "mean_token_accuracy": 0.8273735642433167, "num_tokens": 63927187.0, "step": 1672 }, { "epoch": 0.21282279608192342, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.385451793670654, "learning_rate": 7.08774904620602e-07, "loss": 0.4604, "mean_token_accuracy": 0.8537458777427673, "num_tokens": 63964088.0, "step": 1673 }, { "epoch": 0.21295000636051392, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.393271446228027, "learning_rate": 7.091988130563797e-07, "loss": 0.5052, "mean_token_accuracy": 0.8372111320495605, "num_tokens": 64004931.0, "step": 1674 }, { "epoch": 0.21307721663910445, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.388598918914795, "learning_rate": 7.096227214921576e-07, "loss": 0.4984, "mean_token_accuracy": 0.8387268781661987, "num_tokens": 64043966.0, "step": 1675 }, { "epoch": 0.21320442691769495, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.409492492675781, "learning_rate": 7.100466299279355e-07, "loss": 0.4819, "mean_token_accuracy": 0.8455331921577454, "num_tokens": 64078886.0, "step": 1676 }, { "epoch": 0.21333163719628545, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 7.438588619232178, "learning_rate": 7.104705383637134e-07, "loss": 0.505, "mean_token_accuracy": 0.8418461084365845, "num_tokens": 64117291.0, "step": 1677 }, { "epoch": 0.21345884747487598, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.385584354400635, "learning_rate": 7.108944467994913e-07, "loss": 0.4624, "mean_token_accuracy": 0.8489711284637451, "num_tokens": 64156369.0, "step": 1678 }, { "epoch": 0.21358605775346648, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.420023441314697, "learning_rate": 7.113183552352692e-07, "loss": 0.5355, "mean_token_accuracy": 0.8240196108818054, "num_tokens": 64193646.0, "step": 1679 }, { "epoch": 0.21371326803205698, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.382813453674316, "learning_rate": 7.11742263671047e-07, "loss": 0.428, "mean_token_accuracy": 0.859225869178772, "num_tokens": 64231933.0, "step": 1680 }, { "epoch": 0.2138404783106475, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.465794086456299, "learning_rate": 7.12166172106825e-07, "loss": 0.5597, "mean_token_accuracy": 0.8182007074356079, "num_tokens": 64267418.0, "step": 1681 }, { "epoch": 0.213967688589238, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.406466484069824, "learning_rate": 7.125900805426027e-07, "loss": 0.4415, "mean_token_accuracy": 0.8552436828613281, "num_tokens": 64302440.0, "step": 1682 }, { "epoch": 0.2140948988678285, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 7.412635803222656, "learning_rate": 7.130139889783806e-07, "loss": 0.4802, "mean_token_accuracy": 0.8449653387069702, "num_tokens": 64338240.0, "step": 1683 }, { "epoch": 0.21422210914641904, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.452730655670166, "learning_rate": 7.134378974141585e-07, "loss": 0.4713, "mean_token_accuracy": 0.8461683988571167, "num_tokens": 64375734.0, "step": 1684 }, { "epoch": 0.21434931942500954, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.377135276794434, "learning_rate": 7.138618058499364e-07, "loss": 0.4651, "mean_token_accuracy": 0.8497240543365479, "num_tokens": 64411401.0, "step": 1685 }, { "epoch": 0.21447652970360004, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.409606456756592, "learning_rate": 7.142857142857143e-07, "loss": 0.4402, "mean_token_accuracy": 0.8574122190475464, "num_tokens": 64454719.0, "step": 1686 }, { "epoch": 0.21460373998219057, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.411637783050537, "learning_rate": 7.147096227214922e-07, "loss": 0.4649, "mean_token_accuracy": 0.8505041599273682, "num_tokens": 64495736.0, "step": 1687 }, { "epoch": 0.21473095026078107, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.455310821533203, "learning_rate": 7.1513353115727e-07, "loss": 0.4451, "mean_token_accuracy": 0.8536474704742432, "num_tokens": 64532843.0, "step": 1688 }, { "epoch": 0.21485816053937157, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.429886341094971, "learning_rate": 7.155574395930479e-07, "loss": 0.4907, "mean_token_accuracy": 0.8375751972198486, "num_tokens": 64568950.0, "step": 1689 }, { "epoch": 0.2149853708179621, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.465658187866211, "learning_rate": 7.159813480288257e-07, "loss": 0.4683, "mean_token_accuracy": 0.8448358774185181, "num_tokens": 64606878.0, "step": 1690 }, { "epoch": 0.2151125810965526, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.43706750869751, "learning_rate": 7.164052564646035e-07, "loss": 0.436, "mean_token_accuracy": 0.8575828075408936, "num_tokens": 64643012.0, "step": 1691 }, { "epoch": 0.21523979137514312, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.408080577850342, "learning_rate": 7.168291649003815e-07, "loss": 0.4568, "mean_token_accuracy": 0.8550676703453064, "num_tokens": 64684444.0, "step": 1692 }, { "epoch": 0.21536700165373363, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.486731052398682, "learning_rate": 7.172530733361593e-07, "loss": 0.5064, "mean_token_accuracy": 0.8380756378173828, "num_tokens": 64722068.0, "step": 1693 }, { "epoch": 0.21549421193232413, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 7.469738960266113, "learning_rate": 7.176769817719373e-07, "loss": 0.4181, "mean_token_accuracy": 0.8688058257102966, "num_tokens": 64761829.0, "step": 1694 }, { "epoch": 0.21562142221091465, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.463730335235596, "learning_rate": 7.181008902077151e-07, "loss": 0.4652, "mean_token_accuracy": 0.8492825627326965, "num_tokens": 64799185.0, "step": 1695 }, { "epoch": 0.21574863248950515, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.507968425750732, "learning_rate": 7.18524798643493e-07, "loss": 0.4941, "mean_token_accuracy": 0.8417351245880127, "num_tokens": 64839306.0, "step": 1696 }, { "epoch": 0.21587584276809565, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.467480659484863, "learning_rate": 7.189487070792708e-07, "loss": 0.47, "mean_token_accuracy": 0.8503994941711426, "num_tokens": 64876011.0, "step": 1697 }, { "epoch": 0.21600305304668618, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 7.425858020782471, "learning_rate": 7.193726155150487e-07, "loss": 0.4687, "mean_token_accuracy": 0.8515501618385315, "num_tokens": 64918510.0, "step": 1698 }, { "epoch": 0.21613026332527668, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.463082790374756, "learning_rate": 7.197965239508265e-07, "loss": 0.4751, "mean_token_accuracy": 0.8461466431617737, "num_tokens": 64957156.0, "step": 1699 }, { "epoch": 0.21625747360386718, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.417587757110596, "learning_rate": 7.202204323866045e-07, "loss": 0.4738, "mean_token_accuracy": 0.8482533693313599, "num_tokens": 64998761.0, "step": 1700 }, { "epoch": 0.2163846838824577, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 7.468844890594482, "learning_rate": 7.206443408223823e-07, "loss": 0.4764, "mean_token_accuracy": 0.844332218170166, "num_tokens": 65038849.0, "step": 1701 }, { "epoch": 0.2165118941610482, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.478774547576904, "learning_rate": 7.210682492581603e-07, "loss": 0.4567, "mean_token_accuracy": 0.8491037487983704, "num_tokens": 65071130.0, "step": 1702 }, { "epoch": 0.2166391044396387, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.469398021697998, "learning_rate": 7.214921576939381e-07, "loss": 0.4984, "mean_token_accuracy": 0.8356653451919556, "num_tokens": 65111619.0, "step": 1703 }, { "epoch": 0.21676631471822924, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.484687328338623, "learning_rate": 7.219160661297159e-07, "loss": 0.4453, "mean_token_accuracy": 0.8528264760971069, "num_tokens": 65147537.0, "step": 1704 }, { "epoch": 0.21689352499681974, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.447570323944092, "learning_rate": 7.223399745654938e-07, "loss": 0.4823, "mean_token_accuracy": 0.8430708646774292, "num_tokens": 65187254.0, "step": 1705 }, { "epoch": 0.21702073527541024, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.5032219886779785, "learning_rate": 7.227638830012717e-07, "loss": 0.4939, "mean_token_accuracy": 0.8370121121406555, "num_tokens": 65226130.0, "step": 1706 }, { "epoch": 0.21714794555400077, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.534266948699951, "learning_rate": 7.231877914370495e-07, "loss": 0.4568, "mean_token_accuracy": 0.8531674146652222, "num_tokens": 65261282.0, "step": 1707 }, { "epoch": 0.21727515583259127, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 7.534280776977539, "learning_rate": 7.236116998728275e-07, "loss": 0.5081, "mean_token_accuracy": 0.8356798887252808, "num_tokens": 65294850.0, "step": 1708 }, { "epoch": 0.21740236611118177, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.456263542175293, "learning_rate": 7.240356083086053e-07, "loss": 0.4011, "mean_token_accuracy": 0.8672046661376953, "num_tokens": 65332323.0, "step": 1709 }, { "epoch": 0.2175295763897723, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 7.463267803192139, "learning_rate": 7.244595167443833e-07, "loss": 0.4942, "mean_token_accuracy": 0.8423060178756714, "num_tokens": 65376033.0, "step": 1710 }, { "epoch": 0.2176567866683628, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.453977584838867, "learning_rate": 7.248834251801611e-07, "loss": 0.4595, "mean_token_accuracy": 0.8528646230697632, "num_tokens": 65417875.0, "step": 1711 }, { "epoch": 0.2177839969469533, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.478468418121338, "learning_rate": 7.253073336159388e-07, "loss": 0.4361, "mean_token_accuracy": 0.8575890064239502, "num_tokens": 65456675.0, "step": 1712 }, { "epoch": 0.21791120722554383, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.51315975189209, "learning_rate": 7.257312420517168e-07, "loss": 0.4278, "mean_token_accuracy": 0.8596893548965454, "num_tokens": 65487699.0, "step": 1713 }, { "epoch": 0.21803841750413433, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.511911869049072, "learning_rate": 7.261551504874946e-07, "loss": 0.4555, "mean_token_accuracy": 0.8489824533462524, "num_tokens": 65526143.0, "step": 1714 }, { "epoch": 0.21816562778272486, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 7.467534065246582, "learning_rate": 7.265790589232725e-07, "loss": 0.4529, "mean_token_accuracy": 0.8536668419837952, "num_tokens": 65562969.0, "step": 1715 }, { "epoch": 0.21829283806131536, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.473896503448486, "learning_rate": 7.270029673590504e-07, "loss": 0.4855, "mean_token_accuracy": 0.845390796661377, "num_tokens": 65605999.0, "step": 1716 }, { "epoch": 0.21842004833990586, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 7.5081658363342285, "learning_rate": 7.274268757948283e-07, "loss": 0.5303, "mean_token_accuracy": 0.8312238454818726, "num_tokens": 65644877.0, "step": 1717 }, { "epoch": 0.2185472586184964, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.534675121307373, "learning_rate": 7.278507842306062e-07, "loss": 0.4508, "mean_token_accuracy": 0.8536909818649292, "num_tokens": 65678823.0, "step": 1718 }, { "epoch": 0.2186744688970869, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.508991718292236, "learning_rate": 7.282746926663841e-07, "loss": 0.531, "mean_token_accuracy": 0.830690324306488, "num_tokens": 65717185.0, "step": 1719 }, { "epoch": 0.2188016791756774, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.531904697418213, "learning_rate": 7.286986011021618e-07, "loss": 0.4512, "mean_token_accuracy": 0.8537598848342896, "num_tokens": 65754422.0, "step": 1720 }, { "epoch": 0.21892888945426792, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.493731498718262, "learning_rate": 7.291225095379398e-07, "loss": 0.4874, "mean_token_accuracy": 0.8441439867019653, "num_tokens": 65799390.0, "step": 1721 }, { "epoch": 0.21905609973285842, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.520503520965576, "learning_rate": 7.295464179737176e-07, "loss": 0.4792, "mean_token_accuracy": 0.8480715155601501, "num_tokens": 65837566.0, "step": 1722 }, { "epoch": 0.21918331001144892, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.54741907119751, "learning_rate": 7.299703264094955e-07, "loss": 0.464, "mean_token_accuracy": 0.8491588830947876, "num_tokens": 65874537.0, "step": 1723 }, { "epoch": 0.21931052029003945, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.506356716156006, "learning_rate": 7.303942348452734e-07, "loss": 0.4905, "mean_token_accuracy": 0.8414344787597656, "num_tokens": 65915616.0, "step": 1724 }, { "epoch": 0.21943773056862995, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 7.508522987365723, "learning_rate": 7.308181432810513e-07, "loss": 0.4394, "mean_token_accuracy": 0.8552429676055908, "num_tokens": 65954263.0, "step": 1725 }, { "epoch": 0.21956494084722045, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.538054466247559, "learning_rate": 7.312420517168292e-07, "loss": 0.4726, "mean_token_accuracy": 0.8447245359420776, "num_tokens": 65992880.0, "step": 1726 }, { "epoch": 0.21969215112581097, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.551283836364746, "learning_rate": 7.31665960152607e-07, "loss": 0.4908, "mean_token_accuracy": 0.8384729623794556, "num_tokens": 66033958.0, "step": 1727 }, { "epoch": 0.21981936140440148, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.530318737030029, "learning_rate": 7.320898685883848e-07, "loss": 0.4619, "mean_token_accuracy": 0.8485041856765747, "num_tokens": 66072448.0, "step": 1728 }, { "epoch": 0.21994657168299198, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.534977912902832, "learning_rate": 7.325137770241628e-07, "loss": 0.5728, "mean_token_accuracy": 0.8189014196395874, "num_tokens": 66112648.0, "step": 1729 }, { "epoch": 0.2200737819615825, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.5514349937438965, "learning_rate": 7.329376854599406e-07, "loss": 0.4839, "mean_token_accuracy": 0.8448059558868408, "num_tokens": 66154791.0, "step": 1730 }, { "epoch": 0.220200992240173, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.5655646324157715, "learning_rate": 7.333615938957184e-07, "loss": 0.4714, "mean_token_accuracy": 0.8467470407485962, "num_tokens": 66192450.0, "step": 1731 }, { "epoch": 0.2203282025187635, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.565644264221191, "learning_rate": 7.337855023314964e-07, "loss": 0.4668, "mean_token_accuracy": 0.8470818996429443, "num_tokens": 66231173.0, "step": 1732 }, { "epoch": 0.22045541279735403, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.554666519165039, "learning_rate": 7.342094107672742e-07, "loss": 0.5199, "mean_token_accuracy": 0.8313915133476257, "num_tokens": 66272956.0, "step": 1733 }, { "epoch": 0.22058262307594453, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.573005199432373, "learning_rate": 7.346333192030522e-07, "loss": 0.4742, "mean_token_accuracy": 0.8511247634887695, "num_tokens": 66309450.0, "step": 1734 }, { "epoch": 0.22070983335453503, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.612321853637695, "learning_rate": 7.350572276388299e-07, "loss": 0.5238, "mean_token_accuracy": 0.8298917412757874, "num_tokens": 66350947.0, "step": 1735 }, { "epoch": 0.22083704363312556, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 7.558862209320068, "learning_rate": 7.354811360746078e-07, "loss": 0.4818, "mean_token_accuracy": 0.8428834676742554, "num_tokens": 66388364.0, "step": 1736 }, { "epoch": 0.22096425391171606, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.576992988586426, "learning_rate": 7.359050445103857e-07, "loss": 0.4962, "mean_token_accuracy": 0.8385006189346313, "num_tokens": 66425029.0, "step": 1737 }, { "epoch": 0.22109146419030656, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.603090286254883, "learning_rate": 7.363289529461636e-07, "loss": 0.4721, "mean_token_accuracy": 0.8470028042793274, "num_tokens": 66463924.0, "step": 1738 }, { "epoch": 0.2212186744688971, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.572257041931152, "learning_rate": 7.367528613819415e-07, "loss": 0.427, "mean_token_accuracy": 0.8598642945289612, "num_tokens": 66504824.0, "step": 1739 }, { "epoch": 0.2213458847474876, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.620985984802246, "learning_rate": 7.371767698177194e-07, "loss": 0.4462, "mean_token_accuracy": 0.8544326424598694, "num_tokens": 66539308.0, "step": 1740 }, { "epoch": 0.22147309502607812, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.635448455810547, "learning_rate": 7.376006782534972e-07, "loss": 0.4422, "mean_token_accuracy": 0.8560057878494263, "num_tokens": 66572785.0, "step": 1741 }, { "epoch": 0.22160030530466862, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.61981201171875, "learning_rate": 7.380245866892751e-07, "loss": 0.4594, "mean_token_accuracy": 0.8506543636322021, "num_tokens": 66607890.0, "step": 1742 }, { "epoch": 0.22172751558325912, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 7.545216083526611, "learning_rate": 7.384484951250529e-07, "loss": 0.4616, "mean_token_accuracy": 0.8519595861434937, "num_tokens": 66650553.0, "step": 1743 }, { "epoch": 0.22185472586184965, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.630210876464844, "learning_rate": 7.388724035608308e-07, "loss": 0.4568, "mean_token_accuracy": 0.8523441553115845, "num_tokens": 66685330.0, "step": 1744 }, { "epoch": 0.22198193614044015, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.618372917175293, "learning_rate": 7.392963119966087e-07, "loss": 0.4401, "mean_token_accuracy": 0.8550679087638855, "num_tokens": 66716816.0, "step": 1745 }, { "epoch": 0.22210914641903065, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.652529716491699, "learning_rate": 7.397202204323866e-07, "loss": 0.4845, "mean_token_accuracy": 0.8444145917892456, "num_tokens": 66754030.0, "step": 1746 }, { "epoch": 0.22223635669762118, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 7.620774269104004, "learning_rate": 7.401441288681645e-07, "loss": 0.442, "mean_token_accuracy": 0.8562251329421997, "num_tokens": 66793650.0, "step": 1747 }, { "epoch": 0.22236356697621168, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.567447662353516, "learning_rate": 7.405680373039424e-07, "loss": 0.426, "mean_token_accuracy": 0.8639380931854248, "num_tokens": 66833709.0, "step": 1748 }, { "epoch": 0.22249077725480218, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 7.626987457275391, "learning_rate": 7.409919457397202e-07, "loss": 0.4936, "mean_token_accuracy": 0.8389550447463989, "num_tokens": 66866456.0, "step": 1749 }, { "epoch": 0.2226179875333927, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.590608596801758, "learning_rate": 7.414158541754981e-07, "loss": 0.5202, "mean_token_accuracy": 0.8310920000076294, "num_tokens": 66905875.0, "step": 1750 }, { "epoch": 0.2227451978119832, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.725536346435547, "learning_rate": 7.418397626112759e-07, "loss": 0.4747, "mean_token_accuracy": 0.8535451292991638, "num_tokens": 66936009.0, "step": 1751 }, { "epoch": 0.2228724080905737, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.597845077514648, "learning_rate": 7.422636710470537e-07, "loss": 0.4786, "mean_token_accuracy": 0.8457426428794861, "num_tokens": 66978111.0, "step": 1752 }, { "epoch": 0.22299961836916424, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.609859943389893, "learning_rate": 7.426875794828317e-07, "loss": 0.4911, "mean_token_accuracy": 0.8443588614463806, "num_tokens": 67015938.0, "step": 1753 }, { "epoch": 0.22312682864775474, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.61224889755249, "learning_rate": 7.431114879186095e-07, "loss": 0.5128, "mean_token_accuracy": 0.8350389003753662, "num_tokens": 67055456.0, "step": 1754 }, { "epoch": 0.22325403892634524, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.639969348907471, "learning_rate": 7.435353963543875e-07, "loss": 0.4703, "mean_token_accuracy": 0.8473016023635864, "num_tokens": 67096298.0, "step": 1755 }, { "epoch": 0.22338124920493577, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.599275588989258, "learning_rate": 7.439593047901653e-07, "loss": 0.4263, "mean_token_accuracy": 0.8610179424285889, "num_tokens": 67135280.0, "step": 1756 }, { "epoch": 0.22350845948352627, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.647242546081543, "learning_rate": 7.443832132259431e-07, "loss": 0.4903, "mean_token_accuracy": 0.8397517800331116, "num_tokens": 67171936.0, "step": 1757 }, { "epoch": 0.22363566976211677, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.602136611938477, "learning_rate": 7.44807121661721e-07, "loss": 0.4786, "mean_token_accuracy": 0.8442786931991577, "num_tokens": 67209854.0, "step": 1758 }, { "epoch": 0.2237628800407073, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.660550594329834, "learning_rate": 7.452310300974989e-07, "loss": 0.4338, "mean_token_accuracy": 0.8536562323570251, "num_tokens": 67244369.0, "step": 1759 }, { "epoch": 0.2238900903192978, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 7.591411113739014, "learning_rate": 7.456549385332767e-07, "loss": 0.5117, "mean_token_accuracy": 0.8356708288192749, "num_tokens": 67283994.0, "step": 1760 }, { "epoch": 0.2240173005978883, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.630220413208008, "learning_rate": 7.460788469690547e-07, "loss": 0.3977, "mean_token_accuracy": 0.8664988279342651, "num_tokens": 67323230.0, "step": 1761 }, { "epoch": 0.22414451087647883, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 7.670346736907959, "learning_rate": 7.465027554048325e-07, "loss": 0.5502, "mean_token_accuracy": 0.825547993183136, "num_tokens": 67366198.0, "step": 1762 }, { "epoch": 0.22427172115506933, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.658288955688477, "learning_rate": 7.469266638406105e-07, "loss": 0.5344, "mean_token_accuracy": 0.8308814167976379, "num_tokens": 67410685.0, "step": 1763 }, { "epoch": 0.22439893143365983, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.687771320343018, "learning_rate": 7.473505722763883e-07, "loss": 0.457, "mean_token_accuracy": 0.8530277609825134, "num_tokens": 67445818.0, "step": 1764 }, { "epoch": 0.22452614171225035, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.663313388824463, "learning_rate": 7.477744807121661e-07, "loss": 0.4989, "mean_token_accuracy": 0.8384734988212585, "num_tokens": 67483873.0, "step": 1765 }, { "epoch": 0.22465335199084085, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.718358039855957, "learning_rate": 7.48198389147944e-07, "loss": 0.4768, "mean_token_accuracy": 0.849441647529602, "num_tokens": 67519524.0, "step": 1766 }, { "epoch": 0.22478056226943138, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.634641647338867, "learning_rate": 7.486222975837219e-07, "loss": 0.4616, "mean_token_accuracy": 0.8488446474075317, "num_tokens": 67560920.0, "step": 1767 }, { "epoch": 0.22490777254802188, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.700024127960205, "learning_rate": 7.490462060194997e-07, "loss": 0.4329, "mean_token_accuracy": 0.8554730415344238, "num_tokens": 67594948.0, "step": 1768 }, { "epoch": 0.22503498282661238, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.6835408210754395, "learning_rate": 7.494701144552777e-07, "loss": 0.5195, "mean_token_accuracy": 0.8325971364974976, "num_tokens": 67628867.0, "step": 1769 }, { "epoch": 0.2251621931052029, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 7.653441905975342, "learning_rate": 7.498940228910555e-07, "loss": 0.4805, "mean_token_accuracy": 0.8431804180145264, "num_tokens": 67669762.0, "step": 1770 }, { "epoch": 0.2252894033837934, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 7.710681915283203, "learning_rate": 7.503179313268335e-07, "loss": 0.4679, "mean_token_accuracy": 0.8462153673171997, "num_tokens": 67704750.0, "step": 1771 }, { "epoch": 0.2254166136623839, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.67262601852417, "learning_rate": 7.507418397626113e-07, "loss": 0.5051, "mean_token_accuracy": 0.8366225957870483, "num_tokens": 67745181.0, "step": 1772 }, { "epoch": 0.22554382394097444, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.720966339111328, "learning_rate": 7.51165748198389e-07, "loss": 0.4925, "mean_token_accuracy": 0.8385370373725891, "num_tokens": 67785929.0, "step": 1773 }, { "epoch": 0.22567103421956494, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 7.687500953674316, "learning_rate": 7.51589656634167e-07, "loss": 0.4482, "mean_token_accuracy": 0.8499900102615356, "num_tokens": 67821887.0, "step": 1774 }, { "epoch": 0.22579824449815544, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.629362106323242, "learning_rate": 7.520135650699448e-07, "loss": 0.4824, "mean_token_accuracy": 0.8431150913238525, "num_tokens": 67859608.0, "step": 1775 }, { "epoch": 0.22592545477674597, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 7.704063892364502, "learning_rate": 7.524374735057227e-07, "loss": 0.4916, "mean_token_accuracy": 0.8418315649032593, "num_tokens": 67895287.0, "step": 1776 }, { "epoch": 0.22605266505533647, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.6781535148620605, "learning_rate": 7.528613819415006e-07, "loss": 0.4505, "mean_token_accuracy": 0.8579167723655701, "num_tokens": 67930358.0, "step": 1777 }, { "epoch": 0.22617987533392697, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 7.672569751739502, "learning_rate": 7.532852903772785e-07, "loss": 0.4907, "mean_token_accuracy": 0.8400226831436157, "num_tokens": 67966112.0, "step": 1778 }, { "epoch": 0.2263070856125175, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.686544895172119, "learning_rate": 7.537091988130564e-07, "loss": 0.4987, "mean_token_accuracy": 0.8357110619544983, "num_tokens": 68001297.0, "step": 1779 }, { "epoch": 0.226434295891108, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 7.68184232711792, "learning_rate": 7.541331072488342e-07, "loss": 0.4797, "mean_token_accuracy": 0.8452045321464539, "num_tokens": 68037346.0, "step": 1780 }, { "epoch": 0.2265615061696985, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.699224472045898, "learning_rate": 7.54557015684612e-07, "loss": 0.4786, "mean_token_accuracy": 0.8436918258666992, "num_tokens": 68076566.0, "step": 1781 }, { "epoch": 0.22668871644828903, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.730720043182373, "learning_rate": 7.5498092412039e-07, "loss": 0.4569, "mean_token_accuracy": 0.8539071679115295, "num_tokens": 68116834.0, "step": 1782 }, { "epoch": 0.22681592672687953, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 7.711886405944824, "learning_rate": 7.554048325561678e-07, "loss": 0.4313, "mean_token_accuracy": 0.8600350618362427, "num_tokens": 68155573.0, "step": 1783 }, { "epoch": 0.22694313700547003, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.69477653503418, "learning_rate": 7.558287409919457e-07, "loss": 0.5244, "mean_token_accuracy": 0.8338969945907593, "num_tokens": 68193378.0, "step": 1784 }, { "epoch": 0.22707034728406056, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.7200422286987305, "learning_rate": 7.562526494277236e-07, "loss": 0.4979, "mean_token_accuracy": 0.8355851173400879, "num_tokens": 68230183.0, "step": 1785 }, { "epoch": 0.22719755756265106, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.71012020111084, "learning_rate": 7.566765578635015e-07, "loss": 0.4668, "mean_token_accuracy": 0.8497097492218018, "num_tokens": 68270526.0, "step": 1786 }, { "epoch": 0.22732476784124156, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.728767395019531, "learning_rate": 7.571004662992794e-07, "loss": 0.4922, "mean_token_accuracy": 0.8449056148529053, "num_tokens": 68308690.0, "step": 1787 }, { "epoch": 0.2274519781198321, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.714095115661621, "learning_rate": 7.575243747350572e-07, "loss": 0.5344, "mean_token_accuracy": 0.8310263156890869, "num_tokens": 68348660.0, "step": 1788 }, { "epoch": 0.2275791883984226, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.758533477783203, "learning_rate": 7.57948283170835e-07, "loss": 0.4832, "mean_token_accuracy": 0.8467473983764648, "num_tokens": 68389129.0, "step": 1789 }, { "epoch": 0.2277063986770131, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.793121814727783, "learning_rate": 7.58372191606613e-07, "loss": 0.4754, "mean_token_accuracy": 0.8476042747497559, "num_tokens": 68423234.0, "step": 1790 }, { "epoch": 0.22783360895560362, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.703677654266357, "learning_rate": 7.587961000423908e-07, "loss": 0.4824, "mean_token_accuracy": 0.8457589149475098, "num_tokens": 68460834.0, "step": 1791 }, { "epoch": 0.22796081923419412, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.842505931854248, "learning_rate": 7.592200084781686e-07, "loss": 0.5266, "mean_token_accuracy": 0.8348342180252075, "num_tokens": 68495605.0, "step": 1792 }, { "epoch": 0.22808802951278465, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.757111549377441, "learning_rate": 7.596439169139466e-07, "loss": 0.4728, "mean_token_accuracy": 0.8473525047302246, "num_tokens": 68530838.0, "step": 1793 }, { "epoch": 0.22821523979137515, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 7.741725921630859, "learning_rate": 7.600678253497244e-07, "loss": 0.467, "mean_token_accuracy": 0.8507322669029236, "num_tokens": 68571225.0, "step": 1794 }, { "epoch": 0.22834245006996565, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.730762004852295, "learning_rate": 7.604917337855023e-07, "loss": 0.4485, "mean_token_accuracy": 0.8563135266304016, "num_tokens": 68608284.0, "step": 1795 }, { "epoch": 0.22846966034855618, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.74301290512085, "learning_rate": 7.609156422212801e-07, "loss": 0.4303, "mean_token_accuracy": 0.8588974475860596, "num_tokens": 68645943.0, "step": 1796 }, { "epoch": 0.22859687062714668, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.737274646759033, "learning_rate": 7.61339550657058e-07, "loss": 0.4533, "mean_token_accuracy": 0.8517765402793884, "num_tokens": 68684245.0, "step": 1797 }, { "epoch": 0.22872408090573718, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.740133285522461, "learning_rate": 7.617634590928359e-07, "loss": 0.4669, "mean_token_accuracy": 0.8518984913825989, "num_tokens": 68717532.0, "step": 1798 }, { "epoch": 0.2288512911843277, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.753255367279053, "learning_rate": 7.621873675286138e-07, "loss": 0.486, "mean_token_accuracy": 0.8448579907417297, "num_tokens": 68759515.0, "step": 1799 }, { "epoch": 0.2289785014629182, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.7403950691223145, "learning_rate": 7.626112759643916e-07, "loss": 0.4921, "mean_token_accuracy": 0.8402514457702637, "num_tokens": 68796969.0, "step": 1800 }, { "epoch": 0.2291057117415087, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.80167818069458, "learning_rate": 7.630351844001696e-07, "loss": 0.4735, "mean_token_accuracy": 0.8461618423461914, "num_tokens": 68831138.0, "step": 1801 }, { "epoch": 0.22923292202009923, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.69757080078125, "learning_rate": 7.634590928359474e-07, "loss": 0.4631, "mean_token_accuracy": 0.8519220352172852, "num_tokens": 68868741.0, "step": 1802 }, { "epoch": 0.22936013229868973, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.7153496742248535, "learning_rate": 7.638830012717253e-07, "loss": 0.4841, "mean_token_accuracy": 0.8420182466506958, "num_tokens": 68909191.0, "step": 1803 }, { "epoch": 0.22948734257728023, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.697531223297119, "learning_rate": 7.643069097075031e-07, "loss": 0.4595, "mean_token_accuracy": 0.8493574261665344, "num_tokens": 68949261.0, "step": 1804 }, { "epoch": 0.22961455285587076, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.710751056671143, "learning_rate": 7.64730818143281e-07, "loss": 0.4335, "mean_token_accuracy": 0.8554196357727051, "num_tokens": 68990434.0, "step": 1805 }, { "epoch": 0.22974176313446126, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.72026252746582, "learning_rate": 7.651547265790589e-07, "loss": 0.516, "mean_token_accuracy": 0.835249662399292, "num_tokens": 69029482.0, "step": 1806 }, { "epoch": 0.22986897341305176, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.731202602386475, "learning_rate": 7.655786350148368e-07, "loss": 0.4961, "mean_token_accuracy": 0.8392601013183594, "num_tokens": 69069766.0, "step": 1807 }, { "epoch": 0.2299961836916423, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.7149152755737305, "learning_rate": 7.660025434506146e-07, "loss": 0.4824, "mean_token_accuracy": 0.8442842960357666, "num_tokens": 69114592.0, "step": 1808 }, { "epoch": 0.2301233939702328, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.757458686828613, "learning_rate": 7.664264518863926e-07, "loss": 0.4716, "mean_token_accuracy": 0.8477487564086914, "num_tokens": 69152599.0, "step": 1809 }, { "epoch": 0.2302506042488233, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.729866027832031, "learning_rate": 7.668503603221704e-07, "loss": 0.5236, "mean_token_accuracy": 0.8288902640342712, "num_tokens": 69193157.0, "step": 1810 }, { "epoch": 0.23037781452741382, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 7.746213436126709, "learning_rate": 7.672742687579483e-07, "loss": 0.4354, "mean_token_accuracy": 0.8565620183944702, "num_tokens": 69230681.0, "step": 1811 }, { "epoch": 0.23050502480600432, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.785152435302734, "learning_rate": 7.676981771937261e-07, "loss": 0.4146, "mean_token_accuracy": 0.8593850135803223, "num_tokens": 69261292.0, "step": 1812 }, { "epoch": 0.23063223508459482, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.812502861022949, "learning_rate": 7.681220856295039e-07, "loss": 0.4453, "mean_token_accuracy": 0.8547123670578003, "num_tokens": 69301795.0, "step": 1813 }, { "epoch": 0.23075944536318535, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.814498424530029, "learning_rate": 7.685459940652819e-07, "loss": 0.4684, "mean_token_accuracy": 0.8450579643249512, "num_tokens": 69337038.0, "step": 1814 }, { "epoch": 0.23088665564177585, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 7.7427754402160645, "learning_rate": 7.689699025010597e-07, "loss": 0.4654, "mean_token_accuracy": 0.8473566770553589, "num_tokens": 69379871.0, "step": 1815 }, { "epoch": 0.23101386592036638, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.78687047958374, "learning_rate": 7.693938109368376e-07, "loss": 0.4866, "mean_token_accuracy": 0.8409364819526672, "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": 7.739887237548828, "learning_rate": 7.698177193726155e-07, "loss": 0.4657, "mean_token_accuracy": 0.8474535942077637, "num_tokens": 69455574.0, "step": 1817 }, { "epoch": 0.23126828647754738, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 7.745565891265869, "learning_rate": 7.702416278083933e-07, "loss": 0.4896, "mean_token_accuracy": 0.8425555229187012, "num_tokens": 69489822.0, "step": 1818 }, { "epoch": 0.2313954967561379, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.773576259613037, "learning_rate": 7.706655362441712e-07, "loss": 0.4603, "mean_token_accuracy": 0.8487744331359863, "num_tokens": 69532821.0, "step": 1819 }, { "epoch": 0.2315227070347284, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.81250524520874, "learning_rate": 7.710894446799491e-07, "loss": 0.4444, "mean_token_accuracy": 0.8531840443611145, "num_tokens": 69565674.0, "step": 1820 }, { "epoch": 0.2316499173133189, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.794965744018555, "learning_rate": 7.715133531157269e-07, "loss": 0.4445, "mean_token_accuracy": 0.8577311635017395, "num_tokens": 69606118.0, "step": 1821 }, { "epoch": 0.23177712759190944, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 7.767915725708008, "learning_rate": 7.719372615515049e-07, "loss": 0.5145, "mean_token_accuracy": 0.8346885442733765, "num_tokens": 69650102.0, "step": 1822 }, { "epoch": 0.23190433787049994, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.735768795013428, "learning_rate": 7.723611699872827e-07, "loss": 0.4596, "mean_token_accuracy": 0.8490769863128662, "num_tokens": 69693440.0, "step": 1823 }, { "epoch": 0.23203154814909044, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 7.758511066436768, "learning_rate": 7.727850784230606e-07, "loss": 0.4793, "mean_token_accuracy": 0.8405177593231201, "num_tokens": 69732651.0, "step": 1824 }, { "epoch": 0.23215875842768097, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.846570014953613, "learning_rate": 7.732089868588385e-07, "loss": 0.4941, "mean_token_accuracy": 0.843252420425415, "num_tokens": 69769915.0, "step": 1825 }, { "epoch": 0.23228596870627147, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.714750289916992, "learning_rate": 7.736328952946163e-07, "loss": 0.4165, "mean_token_accuracy": 0.8643525838851929, "num_tokens": 69813943.0, "step": 1826 }, { "epoch": 0.23241317898486197, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 7.886536598205566, "learning_rate": 7.740568037303942e-07, "loss": 0.4462, "mean_token_accuracy": 0.8518770337104797, "num_tokens": 69847039.0, "step": 1827 }, { "epoch": 0.2325403892634525, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.819102764129639, "learning_rate": 7.744807121661721e-07, "loss": 0.4114, "mean_token_accuracy": 0.867897629737854, "num_tokens": 69884588.0, "step": 1828 }, { "epoch": 0.232667599542043, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.800805568695068, "learning_rate": 7.749046206019499e-07, "loss": 0.4753, "mean_token_accuracy": 0.8461519479751587, "num_tokens": 69925552.0, "step": 1829 }, { "epoch": 0.2327948098206335, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.77512788772583, "learning_rate": 7.753285290377279e-07, "loss": 0.4093, "mean_token_accuracy": 0.8646712303161621, "num_tokens": 69961776.0, "step": 1830 }, { "epoch": 0.23292202009922403, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.849323749542236, "learning_rate": 7.757524374735057e-07, "loss": 0.5156, "mean_token_accuracy": 0.8330389261245728, "num_tokens": 69999775.0, "step": 1831 }, { "epoch": 0.23304923037781453, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.822712421417236, "learning_rate": 7.761763459092836e-07, "loss": 0.426, "mean_token_accuracy": 0.8627232313156128, "num_tokens": 70037294.0, "step": 1832 }, { "epoch": 0.23317644065640503, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.793889045715332, "learning_rate": 7.766002543450614e-07, "loss": 0.4619, "mean_token_accuracy": 0.8509234189987183, "num_tokens": 70075014.0, "step": 1833 }, { "epoch": 0.23330365093499555, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.795549392700195, "learning_rate": 7.770241627808392e-07, "loss": 0.4184, "mean_token_accuracy": 0.8627789616584778, "num_tokens": 70112186.0, "step": 1834 }, { "epoch": 0.23343086121358606, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 7.795076847076416, "learning_rate": 7.774480712166172e-07, "loss": 0.4959, "mean_token_accuracy": 0.8391528129577637, "num_tokens": 70153285.0, "step": 1835 }, { "epoch": 0.23355807149217656, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.843563079833984, "learning_rate": 7.77871979652395e-07, "loss": 0.4786, "mean_token_accuracy": 0.8448492884635925, "num_tokens": 70193382.0, "step": 1836 }, { "epoch": 0.23368528177076708, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.845798015594482, "learning_rate": 7.782958880881729e-07, "loss": 0.4681, "mean_token_accuracy": 0.8479058742523193, "num_tokens": 70227279.0, "step": 1837 }, { "epoch": 0.23381249204935758, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.83718204498291, "learning_rate": 7.787197965239508e-07, "loss": 0.4305, "mean_token_accuracy": 0.8560067415237427, "num_tokens": 70264395.0, "step": 1838 }, { "epoch": 0.23393970232794808, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.933986663818359, "learning_rate": 7.791437049597287e-07, "loss": 0.5456, "mean_token_accuracy": 0.825449526309967, "num_tokens": 70295016.0, "step": 1839 }, { "epoch": 0.2340669126065386, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.812105178833008, "learning_rate": 7.795676133955065e-07, "loss": 0.4891, "mean_token_accuracy": 0.8428421020507812, "num_tokens": 70332774.0, "step": 1840 }, { "epoch": 0.2341941228851291, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 7.778143405914307, "learning_rate": 7.799915218312844e-07, "loss": 0.4745, "mean_token_accuracy": 0.8455819487571716, "num_tokens": 70366880.0, "step": 1841 }, { "epoch": 0.23432133316371964, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.8475728034973145, "learning_rate": 7.804154302670622e-07, "loss": 0.45, "mean_token_accuracy": 0.8542222380638123, "num_tokens": 70405235.0, "step": 1842 }, { "epoch": 0.23444854344231014, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.818458080291748, "learning_rate": 7.808393387028402e-07, "loss": 0.4591, "mean_token_accuracy": 0.8490638732910156, "num_tokens": 70446503.0, "step": 1843 }, { "epoch": 0.23457575372090064, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 7.819579124450684, "learning_rate": 7.81263247138618e-07, "loss": 0.5101, "mean_token_accuracy": 0.8345906138420105, "num_tokens": 70483514.0, "step": 1844 }, { "epoch": 0.23470296399949117, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.821876049041748, "learning_rate": 7.816871555743959e-07, "loss": 0.4712, "mean_token_accuracy": 0.8460619449615479, "num_tokens": 70520944.0, "step": 1845 }, { "epoch": 0.23483017427808167, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.8550944328308105, "learning_rate": 7.821110640101738e-07, "loss": 0.4946, "mean_token_accuracy": 0.8372213244438171, "num_tokens": 70564399.0, "step": 1846 }, { "epoch": 0.23495738455667217, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.836787700653076, "learning_rate": 7.825349724459517e-07, "loss": 0.4679, "mean_token_accuracy": 0.8484160304069519, "num_tokens": 70604000.0, "step": 1847 }, { "epoch": 0.2350845948352627, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.820554256439209, "learning_rate": 7.829588808817294e-07, "loss": 0.4608, "mean_token_accuracy": 0.8494496941566467, "num_tokens": 70645133.0, "step": 1848 }, { "epoch": 0.2352118051138532, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.910068511962891, "learning_rate": 7.833827893175074e-07, "loss": 0.4949, "mean_token_accuracy": 0.8341556787490845, "num_tokens": 70675824.0, "step": 1849 }, { "epoch": 0.2353390153924437, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.804315090179443, "learning_rate": 7.838066977532852e-07, "loss": 0.4579, "mean_token_accuracy": 0.8503544926643372, "num_tokens": 70714103.0, "step": 1850 }, { "epoch": 0.23546622567103423, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 7.830536842346191, "learning_rate": 7.842306061890632e-07, "loss": 0.507, "mean_token_accuracy": 0.8379080295562744, "num_tokens": 70754265.0, "step": 1851 }, { "epoch": 0.23559343594962473, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 7.849029064178467, "learning_rate": 7.84654514624841e-07, "loss": 0.475, "mean_token_accuracy": 0.8429539203643799, "num_tokens": 70791368.0, "step": 1852 }, { "epoch": 0.23572064622821523, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.900242328643799, "learning_rate": 7.850784230606188e-07, "loss": 0.5197, "mean_token_accuracy": 0.8357009291648865, "num_tokens": 70827397.0, "step": 1853 }, { "epoch": 0.23584785650680576, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.813470840454102, "learning_rate": 7.855023314963968e-07, "loss": 0.4903, "mean_token_accuracy": 0.8434854745864868, "num_tokens": 70872503.0, "step": 1854 }, { "epoch": 0.23597506678539626, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.88181209564209, "learning_rate": 7.859262399321746e-07, "loss": 0.49, "mean_token_accuracy": 0.8401831388473511, "num_tokens": 70914376.0, "step": 1855 }, { "epoch": 0.23610227706398676, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.875568866729736, "learning_rate": 7.863501483679524e-07, "loss": 0.447, "mean_token_accuracy": 0.8548062443733215, "num_tokens": 70954004.0, "step": 1856 }, { "epoch": 0.2362294873425773, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.911407470703125, "learning_rate": 7.867740568037303e-07, "loss": 0.4473, "mean_token_accuracy": 0.8511173725128174, "num_tokens": 70991765.0, "step": 1857 }, { "epoch": 0.2363566976211678, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.892495155334473, "learning_rate": 7.871979652395082e-07, "loss": 0.4763, "mean_token_accuracy": 0.8419933915138245, "num_tokens": 71025478.0, "step": 1858 }, { "epoch": 0.2364839078997583, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.838601112365723, "learning_rate": 7.876218736752861e-07, "loss": 0.431, "mean_token_accuracy": 0.8614887595176697, "num_tokens": 71067287.0, "step": 1859 }, { "epoch": 0.23661111817834882, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.880327224731445, "learning_rate": 7.88045782111064e-07, "loss": 0.5003, "mean_token_accuracy": 0.8396766185760498, "num_tokens": 71108095.0, "step": 1860 }, { "epoch": 0.23673832845693932, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.87313985824585, "learning_rate": 7.884696905468418e-07, "loss": 0.4346, "mean_token_accuracy": 0.8593055605888367, "num_tokens": 71150297.0, "step": 1861 }, { "epoch": 0.23686553873552982, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.8713507652282715, "learning_rate": 7.888935989826198e-07, "loss": 0.4823, "mean_token_accuracy": 0.8414818644523621, "num_tokens": 71189205.0, "step": 1862 }, { "epoch": 0.23699274901412035, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.867973327636719, "learning_rate": 7.893175074183976e-07, "loss": 0.4716, "mean_token_accuracy": 0.8468888998031616, "num_tokens": 71228141.0, "step": 1863 }, { "epoch": 0.23711995929271085, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.8877272605896, "learning_rate": 7.897414158541754e-07, "loss": 0.4711, "mean_token_accuracy": 0.846861720085144, "num_tokens": 71267854.0, "step": 1864 }, { "epoch": 0.23724716957130135, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.94608211517334, "learning_rate": 7.901653242899533e-07, "loss": 0.5069, "mean_token_accuracy": 0.8405508399009705, "num_tokens": 71305030.0, "step": 1865 }, { "epoch": 0.23737437984989188, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.8885722160339355, "learning_rate": 7.905892327257312e-07, "loss": 0.4625, "mean_token_accuracy": 0.849247932434082, "num_tokens": 71342202.0, "step": 1866 }, { "epoch": 0.23750159012848238, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 7.911891460418701, "learning_rate": 7.910131411615091e-07, "loss": 0.4627, "mean_token_accuracy": 0.8517199754714966, "num_tokens": 71379702.0, "step": 1867 }, { "epoch": 0.2376288004070729, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.890413761138916, "learning_rate": 7.91437049597287e-07, "loss": 0.4644, "mean_token_accuracy": 0.8489735126495361, "num_tokens": 71417785.0, "step": 1868 }, { "epoch": 0.2377560106856634, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.824199199676514, "learning_rate": 7.918609580330648e-07, "loss": 0.5083, "mean_token_accuracy": 0.8355274200439453, "num_tokens": 71461358.0, "step": 1869 }, { "epoch": 0.2378832209642539, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.915622711181641, "learning_rate": 7.922848664688428e-07, "loss": 0.5057, "mean_token_accuracy": 0.8388149738311768, "num_tokens": 71500059.0, "step": 1870 }, { "epoch": 0.23801043124284443, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.909678936004639, "learning_rate": 7.927087749046205e-07, "loss": 0.4165, "mean_token_accuracy": 0.8646336793899536, "num_tokens": 71537043.0, "step": 1871 }, { "epoch": 0.23813764152143493, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.900359153747559, "learning_rate": 7.931326833403983e-07, "loss": 0.4519, "mean_token_accuracy": 0.850742757320404, "num_tokens": 71570671.0, "step": 1872 }, { "epoch": 0.23826485180002543, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.93027925491333, "learning_rate": 7.935565917761763e-07, "loss": 0.4616, "mean_token_accuracy": 0.8507072329521179, "num_tokens": 71614813.0, "step": 1873 }, { "epoch": 0.23839206207861596, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.934113502502441, "learning_rate": 7.939805002119541e-07, "loss": 0.477, "mean_token_accuracy": 0.8438292145729065, "num_tokens": 71650441.0, "step": 1874 }, { "epoch": 0.23851927235720646, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.876120567321777, "learning_rate": 7.944044086477321e-07, "loss": 0.483, "mean_token_accuracy": 0.844720184803009, "num_tokens": 71690425.0, "step": 1875 }, { "epoch": 0.23864648263579696, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 7.862537384033203, "learning_rate": 7.948283170835099e-07, "loss": 0.4306, "mean_token_accuracy": 0.8575382232666016, "num_tokens": 71727349.0, "step": 1876 }, { "epoch": 0.2387736929143875, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.857134819030762, "learning_rate": 7.952522255192878e-07, "loss": 0.3972, "mean_token_accuracy": 0.8694483637809753, "num_tokens": 71770386.0, "step": 1877 }, { "epoch": 0.238900903192978, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.930224895477295, "learning_rate": 7.956761339550657e-07, "loss": 0.4285, "mean_token_accuracy": 0.8597697615623474, "num_tokens": 71808946.0, "step": 1878 }, { "epoch": 0.2390281134715685, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.894017219543457, "learning_rate": 7.961000423908435e-07, "loss": 0.4698, "mean_token_accuracy": 0.8534722328186035, "num_tokens": 71848605.0, "step": 1879 }, { "epoch": 0.23915532375015902, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.919361114501953, "learning_rate": 7.965239508266214e-07, "loss": 0.4484, "mean_token_accuracy": 0.8525747060775757, "num_tokens": 71886771.0, "step": 1880 }, { "epoch": 0.23928253402874952, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.941464424133301, "learning_rate": 7.969478592623993e-07, "loss": 0.4259, "mean_token_accuracy": 0.857659101486206, "num_tokens": 71924905.0, "step": 1881 }, { "epoch": 0.23940974430734002, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.959778785705566, "learning_rate": 7.973717676981771e-07, "loss": 0.4526, "mean_token_accuracy": 0.8537617325782776, "num_tokens": 71964513.0, "step": 1882 }, { "epoch": 0.23953695458593055, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.954269886016846, "learning_rate": 7.977956761339551e-07, "loss": 0.4227, "mean_token_accuracy": 0.8613598346710205, "num_tokens": 72000918.0, "step": 1883 }, { "epoch": 0.23966416486452105, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.95466423034668, "learning_rate": 7.982195845697329e-07, "loss": 0.5525, "mean_token_accuracy": 0.8223076462745667, "num_tokens": 72040840.0, "step": 1884 }, { "epoch": 0.23979137514311155, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.9130144119262695, "learning_rate": 7.986434930055108e-07, "loss": 0.4258, "mean_token_accuracy": 0.8586831092834473, "num_tokens": 72076757.0, "step": 1885 }, { "epoch": 0.23991858542170208, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.922271728515625, "learning_rate": 7.990674014412886e-07, "loss": 0.4819, "mean_token_accuracy": 0.8455559015274048, "num_tokens": 72120760.0, "step": 1886 }, { "epoch": 0.24004579570029258, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.9584760665893555, "learning_rate": 7.994913098770665e-07, "loss": 0.4693, "mean_token_accuracy": 0.8463869094848633, "num_tokens": 72159985.0, "step": 1887 }, { "epoch": 0.24017300597888308, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.938065052032471, "learning_rate": 7.999152183128444e-07, "loss": 0.5024, "mean_token_accuracy": 0.836281418800354, "num_tokens": 72196377.0, "step": 1888 }, { "epoch": 0.2403002162574736, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.95646333694458, "learning_rate": 8.003391267486223e-07, "loss": 0.4505, "mean_token_accuracy": 0.8533263206481934, "num_tokens": 72236585.0, "step": 1889 }, { "epoch": 0.2404274265360641, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.965759754180908, "learning_rate": 8.007630351844001e-07, "loss": 0.463, "mean_token_accuracy": 0.8483033180236816, "num_tokens": 72268450.0, "step": 1890 }, { "epoch": 0.24055463681465464, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 7.947237968444824, "learning_rate": 8.011869436201781e-07, "loss": 0.4473, "mean_token_accuracy": 0.8537765741348267, "num_tokens": 72304092.0, "step": 1891 }, { "epoch": 0.24068184709324514, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.892460823059082, "learning_rate": 8.016108520559559e-07, "loss": 0.474, "mean_token_accuracy": 0.8449045419692993, "num_tokens": 72347312.0, "step": 1892 }, { "epoch": 0.24080905737183564, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.995998859405518, "learning_rate": 8.020347604917338e-07, "loss": 0.4374, "mean_token_accuracy": 0.8566397428512573, "num_tokens": 72382135.0, "step": 1893 }, { "epoch": 0.24093626765042617, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.91457462310791, "learning_rate": 8.024586689275116e-07, "loss": 0.4635, "mean_token_accuracy": 0.8490563631057739, "num_tokens": 72421526.0, "step": 1894 }, { "epoch": 0.24106347792901667, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.982071399688721, "learning_rate": 8.028825773632894e-07, "loss": 0.4651, "mean_token_accuracy": 0.8504254817962646, "num_tokens": 72458592.0, "step": 1895 }, { "epoch": 0.24119068820760717, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.003666877746582, "learning_rate": 8.033064857990674e-07, "loss": 0.3969, "mean_token_accuracy": 0.868436872959137, "num_tokens": 72494734.0, "step": 1896 }, { "epoch": 0.2413178984861977, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.972256660461426, "learning_rate": 8.037303942348452e-07, "loss": 0.4462, "mean_token_accuracy": 0.855126142501831, "num_tokens": 72531560.0, "step": 1897 }, { "epoch": 0.2414451087647882, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.95866584777832, "learning_rate": 8.041543026706231e-07, "loss": 0.4441, "mean_token_accuracy": 0.8556244969367981, "num_tokens": 72570144.0, "step": 1898 }, { "epoch": 0.2415723190433787, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.015735626220703, "learning_rate": 8.04578211106401e-07, "loss": 0.4533, "mean_token_accuracy": 0.8496999740600586, "num_tokens": 72606738.0, "step": 1899 }, { "epoch": 0.24169952932196923, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.968338966369629, "learning_rate": 8.050021195421789e-07, "loss": 0.4969, "mean_token_accuracy": 0.8393597602844238, "num_tokens": 72646230.0, "step": 1900 }, { "epoch": 0.24182673960055973, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.953644275665283, "learning_rate": 8.054260279779567e-07, "loss": 0.4658, "mean_token_accuracy": 0.8455785512924194, "num_tokens": 72681811.0, "step": 1901 }, { "epoch": 0.24195394987915023, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.013736724853516, "learning_rate": 8.058499364137346e-07, "loss": 0.5022, "mean_token_accuracy": 0.837698757648468, "num_tokens": 72721805.0, "step": 1902 }, { "epoch": 0.24208116015774075, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.963737487792969, "learning_rate": 8.062738448495124e-07, "loss": 0.4159, "mean_token_accuracy": 0.8630043268203735, "num_tokens": 72763142.0, "step": 1903 }, { "epoch": 0.24220837043633126, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.996596336364746, "learning_rate": 8.066977532852904e-07, "loss": 0.4071, "mean_token_accuracy": 0.870338499546051, "num_tokens": 72802047.0, "step": 1904 }, { "epoch": 0.24233558071492176, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.016630172729492, "learning_rate": 8.071216617210682e-07, "loss": 0.4849, "mean_token_accuracy": 0.8429250717163086, "num_tokens": 72834470.0, "step": 1905 }, { "epoch": 0.24246279099351228, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.987051486968994, "learning_rate": 8.075455701568461e-07, "loss": 0.4543, "mean_token_accuracy": 0.8526303768157959, "num_tokens": 72873101.0, "step": 1906 }, { "epoch": 0.24259000127210278, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.993609428405762, "learning_rate": 8.07969478592624e-07, "loss": 0.4352, "mean_token_accuracy": 0.860974907875061, "num_tokens": 72915504.0, "step": 1907 }, { "epoch": 0.24271721155069328, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.005655288696289, "learning_rate": 8.083933870284019e-07, "loss": 0.4091, "mean_token_accuracy": 0.8673076033592224, "num_tokens": 72947179.0, "step": 1908 }, { "epoch": 0.2428444218292838, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.015201568603516, "learning_rate": 8.088172954641796e-07, "loss": 0.4493, "mean_token_accuracy": 0.8538503646850586, "num_tokens": 72983313.0, "step": 1909 }, { "epoch": 0.2429716321078743, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.055960655212402, "learning_rate": 8.092412038999576e-07, "loss": 0.4988, "mean_token_accuracy": 0.8396281599998474, "num_tokens": 73014218.0, "step": 1910 }, { "epoch": 0.24309884238646481, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.025185585021973, "learning_rate": 8.096651123357354e-07, "loss": 0.508, "mean_token_accuracy": 0.832658052444458, "num_tokens": 73048282.0, "step": 1911 }, { "epoch": 0.24322605266505534, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.0542631149292, "learning_rate": 8.100890207715134e-07, "loss": 0.4946, "mean_token_accuracy": 0.8405193090438843, "num_tokens": 73084390.0, "step": 1912 }, { "epoch": 0.24335326294364584, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.988953590393066, "learning_rate": 8.105129292072912e-07, "loss": 0.4623, "mean_token_accuracy": 0.8497259020805359, "num_tokens": 73123296.0, "step": 1913 }, { "epoch": 0.24348047322223634, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 8.029989242553711, "learning_rate": 8.10936837643069e-07, "loss": 0.5065, "mean_token_accuracy": 0.8378669023513794, "num_tokens": 73157533.0, "step": 1914 }, { "epoch": 0.24360768350082687, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.998970985412598, "learning_rate": 8.11360746078847e-07, "loss": 0.441, "mean_token_accuracy": 0.8549709916114807, "num_tokens": 73195314.0, "step": 1915 }, { "epoch": 0.24373489377941737, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 7.96409797668457, "learning_rate": 8.117846545146248e-07, "loss": 0.4095, "mean_token_accuracy": 0.8631576895713806, "num_tokens": 73236763.0, "step": 1916 }, { "epoch": 0.2438621040580079, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.037208557128906, "learning_rate": 8.122085629504026e-07, "loss": 0.4975, "mean_token_accuracy": 0.8366215825080872, "num_tokens": 73276719.0, "step": 1917 }, { "epoch": 0.2439893143365984, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.006529808044434, "learning_rate": 8.126324713861805e-07, "loss": 0.4636, "mean_token_accuracy": 0.8517266511917114, "num_tokens": 73318594.0, "step": 1918 }, { "epoch": 0.2441165246151889, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.992921352386475, "learning_rate": 8.130563798219584e-07, "loss": 0.4391, "mean_token_accuracy": 0.8585875034332275, "num_tokens": 73358760.0, "step": 1919 }, { "epoch": 0.24424373489377943, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.011232376098633, "learning_rate": 8.134802882577363e-07, "loss": 0.4819, "mean_token_accuracy": 0.8472187519073486, "num_tokens": 73398623.0, "step": 1920 }, { "epoch": 0.24437094517236993, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.9981184005737305, "learning_rate": 8.139041966935142e-07, "loss": 0.4641, "mean_token_accuracy": 0.8507261276245117, "num_tokens": 73446369.0, "step": 1921 }, { "epoch": 0.24449815545096043, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.038941383361816, "learning_rate": 8.14328105129292e-07, "loss": 0.5237, "mean_token_accuracy": 0.8360495567321777, "num_tokens": 73483152.0, "step": 1922 }, { "epoch": 0.24462536572955096, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.033784866333008, "learning_rate": 8.1475201356507e-07, "loss": 0.4781, "mean_token_accuracy": 0.8454551696777344, "num_tokens": 73519179.0, "step": 1923 }, { "epoch": 0.24475257600814146, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.017898559570312, "learning_rate": 8.151759220008477e-07, "loss": 0.4499, "mean_token_accuracy": 0.8521463871002197, "num_tokens": 73561947.0, "step": 1924 }, { "epoch": 0.24487978628673196, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.004830360412598, "learning_rate": 8.155998304366256e-07, "loss": 0.4648, "mean_token_accuracy": 0.8533316254615784, "num_tokens": 73595841.0, "step": 1925 }, { "epoch": 0.2450069965653225, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.046257972717285, "learning_rate": 8.160237388724035e-07, "loss": 0.4914, "mean_token_accuracy": 0.8385413289070129, "num_tokens": 73629716.0, "step": 1926 }, { "epoch": 0.245134206843913, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.026500701904297, "learning_rate": 8.164476473081814e-07, "loss": 0.4783, "mean_token_accuracy": 0.8453790545463562, "num_tokens": 73667285.0, "step": 1927 }, { "epoch": 0.2452614171225035, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.007316589355469, "learning_rate": 8.168715557439593e-07, "loss": 0.4225, "mean_token_accuracy": 0.8623374700546265, "num_tokens": 73708059.0, "step": 1928 }, { "epoch": 0.24538862740109402, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 7.993751525878906, "learning_rate": 8.172954641797372e-07, "loss": 0.435, "mean_token_accuracy": 0.8567636013031006, "num_tokens": 73748553.0, "step": 1929 }, { "epoch": 0.24551583767968452, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.019993782043457, "learning_rate": 8.17719372615515e-07, "loss": 0.5352, "mean_token_accuracy": 0.8276149034500122, "num_tokens": 73796295.0, "step": 1930 }, { "epoch": 0.24564304795827502, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.036181449890137, "learning_rate": 8.18143281051293e-07, "loss": 0.5148, "mean_token_accuracy": 0.8357856869697571, "num_tokens": 73835960.0, "step": 1931 }, { "epoch": 0.24577025823686555, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.07852554321289, "learning_rate": 8.185671894870707e-07, "loss": 0.5305, "mean_token_accuracy": 0.8299663662910461, "num_tokens": 73875302.0, "step": 1932 }, { "epoch": 0.24589746851545605, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.073802947998047, "learning_rate": 8.189910979228485e-07, "loss": 0.422, "mean_token_accuracy": 0.8608354330062866, "num_tokens": 73907703.0, "step": 1933 }, { "epoch": 0.24602467879404655, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.039870262145996, "learning_rate": 8.194150063586265e-07, "loss": 0.5162, "mean_token_accuracy": 0.8361186981201172, "num_tokens": 73945449.0, "step": 1934 }, { "epoch": 0.24615188907263708, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.08649730682373, "learning_rate": 8.198389147944043e-07, "loss": 0.5037, "mean_token_accuracy": 0.8344676494598389, "num_tokens": 73982461.0, "step": 1935 }, { "epoch": 0.24627909935122758, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.012059211730957, "learning_rate": 8.202628232301823e-07, "loss": 0.477, "mean_token_accuracy": 0.849717915058136, "num_tokens": 74018252.0, "step": 1936 }, { "epoch": 0.24640630962981808, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.048357009887695, "learning_rate": 8.206867316659601e-07, "loss": 0.4979, "mean_token_accuracy": 0.837397575378418, "num_tokens": 74065843.0, "step": 1937 }, { "epoch": 0.2465335199084086, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.020724296569824, "learning_rate": 8.21110640101738e-07, "loss": 0.3887, "mean_token_accuracy": 0.8689361810684204, "num_tokens": 74106389.0, "step": 1938 }, { "epoch": 0.2466607301869991, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.073014259338379, "learning_rate": 8.215345485375159e-07, "loss": 0.5085, "mean_token_accuracy": 0.836335301399231, "num_tokens": 74150786.0, "step": 1939 }, { "epoch": 0.2467879404655896, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.048885345458984, "learning_rate": 8.219584569732937e-07, "loss": 0.4786, "mean_token_accuracy": 0.8417583703994751, "num_tokens": 74190474.0, "step": 1940 }, { "epoch": 0.24691515074418013, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.0629243850708, "learning_rate": 8.223823654090715e-07, "loss": 0.429, "mean_token_accuracy": 0.8599812984466553, "num_tokens": 74226437.0, "step": 1941 }, { "epoch": 0.24704236102277063, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.07122802734375, "learning_rate": 8.228062738448495e-07, "loss": 0.4568, "mean_token_accuracy": 0.8482420444488525, "num_tokens": 74267292.0, "step": 1942 }, { "epoch": 0.24716957130136116, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.05709171295166, "learning_rate": 8.232301822806273e-07, "loss": 0.4825, "mean_token_accuracy": 0.8441812992095947, "num_tokens": 74306738.0, "step": 1943 }, { "epoch": 0.24729678157995166, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.129637718200684, "learning_rate": 8.236540907164053e-07, "loss": 0.4712, "mean_token_accuracy": 0.8495043516159058, "num_tokens": 74343151.0, "step": 1944 }, { "epoch": 0.24742399185854216, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.074288368225098, "learning_rate": 8.240779991521831e-07, "loss": 0.4484, "mean_token_accuracy": 0.8529448509216309, "num_tokens": 74382755.0, "step": 1945 }, { "epoch": 0.2475512021371327, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.067225456237793, "learning_rate": 8.24501907587961e-07, "loss": 0.4846, "mean_token_accuracy": 0.8428912162780762, "num_tokens": 74426129.0, "step": 1946 }, { "epoch": 0.2476784124157232, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.107383728027344, "learning_rate": 8.249258160237388e-07, "loss": 0.496, "mean_token_accuracy": 0.8450287580490112, "num_tokens": 74460904.0, "step": 1947 }, { "epoch": 0.2478056226943137, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.111353874206543, "learning_rate": 8.253497244595167e-07, "loss": 0.4848, "mean_token_accuracy": 0.8431499004364014, "num_tokens": 74497993.0, "step": 1948 }, { "epoch": 0.24793283297290422, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.105745315551758, "learning_rate": 8.257736328952945e-07, "loss": 0.5036, "mean_token_accuracy": 0.8383771181106567, "num_tokens": 74541108.0, "step": 1949 }, { "epoch": 0.24806004325149472, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 8.082462310791016, "learning_rate": 8.261975413310725e-07, "loss": 0.4732, "mean_token_accuracy": 0.8477212190628052, "num_tokens": 74579389.0, "step": 1950 }, { "epoch": 0.24818725353008522, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.076969146728516, "learning_rate": 8.266214497668503e-07, "loss": 0.4269, "mean_token_accuracy": 0.8625496625900269, "num_tokens": 74621546.0, "step": 1951 }, { "epoch": 0.24831446380867575, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.117542266845703, "learning_rate": 8.270453582026283e-07, "loss": 0.5066, "mean_token_accuracy": 0.834557294845581, "num_tokens": 74660178.0, "step": 1952 }, { "epoch": 0.24844167408726625, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.116776466369629, "learning_rate": 8.274692666384061e-07, "loss": 0.4529, "mean_token_accuracy": 0.8525208234786987, "num_tokens": 74693867.0, "step": 1953 }, { "epoch": 0.24856888436585675, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.114344596862793, "learning_rate": 8.27893175074184e-07, "loss": 0.4663, "mean_token_accuracy": 0.8477702736854553, "num_tokens": 74727561.0, "step": 1954 }, { "epoch": 0.24869609464444728, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.137463569641113, "learning_rate": 8.283170835099618e-07, "loss": 0.4689, "mean_token_accuracy": 0.8489457368850708, "num_tokens": 74763719.0, "step": 1955 }, { "epoch": 0.24882330492303778, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.10374641418457, "learning_rate": 8.287409919457396e-07, "loss": 0.4331, "mean_token_accuracy": 0.8573585748672485, "num_tokens": 74802666.0, "step": 1956 }, { "epoch": 0.24895051520162828, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.118805885314941, "learning_rate": 8.291649003815175e-07, "loss": 0.4837, "mean_token_accuracy": 0.8441764116287231, "num_tokens": 74840307.0, "step": 1957 }, { "epoch": 0.2490777254802188, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.098246574401855, "learning_rate": 8.295888088172954e-07, "loss": 0.495, "mean_token_accuracy": 0.8423528075218201, "num_tokens": 74880181.0, "step": 1958 }, { "epoch": 0.2492049357588093, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.099604606628418, "learning_rate": 8.300127172530733e-07, "loss": 0.4033, "mean_token_accuracy": 0.8656694889068604, "num_tokens": 74917148.0, "step": 1959 }, { "epoch": 0.2493321460373998, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.16893196105957, "learning_rate": 8.304366256888512e-07, "loss": 0.4404, "mean_token_accuracy": 0.8551183342933655, "num_tokens": 74953124.0, "step": 1960 }, { "epoch": 0.24945935631599034, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.111676216125488, "learning_rate": 8.308605341246291e-07, "loss": 0.4536, "mean_token_accuracy": 0.8517968654632568, "num_tokens": 74989726.0, "step": 1961 }, { "epoch": 0.24958656659458084, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.099979400634766, "learning_rate": 8.312844425604068e-07, "loss": 0.4555, "mean_token_accuracy": 0.857811689376831, "num_tokens": 75024849.0, "step": 1962 }, { "epoch": 0.24971377687317134, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.1327543258667, "learning_rate": 8.317083509961848e-07, "loss": 0.4825, "mean_token_accuracy": 0.8460971117019653, "num_tokens": 75066464.0, "step": 1963 }, { "epoch": 0.24984098715176187, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.111854553222656, "learning_rate": 8.321322594319626e-07, "loss": 0.4344, "mean_token_accuracy": 0.8566651344299316, "num_tokens": 75106829.0, "step": 1964 }, { "epoch": 0.24996819743035237, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.114727973937988, "learning_rate": 8.325561678677405e-07, "loss": 0.4882, "mean_token_accuracy": 0.8426392078399658, "num_tokens": 75148227.0, "step": 1965 }, { "epoch": 0.2500954077089429, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.141780853271484, "learning_rate": 8.329800763035184e-07, "loss": 0.5014, "mean_token_accuracy": 0.8364283442497253, "num_tokens": 75188719.0, "step": 1966 }, { "epoch": 0.25022261798753337, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.135775566101074, "learning_rate": 8.334039847392963e-07, "loss": 0.4521, "mean_token_accuracy": 0.8506363034248352, "num_tokens": 75222260.0, "step": 1967 }, { "epoch": 0.2503498282661239, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.156188011169434, "learning_rate": 8.338278931750742e-07, "loss": 0.4272, "mean_token_accuracy": 0.8607982397079468, "num_tokens": 75255489.0, "step": 1968 }, { "epoch": 0.2504770385447144, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.120745658874512, "learning_rate": 8.342518016108521e-07, "loss": 0.4533, "mean_token_accuracy": 0.8485274314880371, "num_tokens": 75290548.0, "step": 1969 }, { "epoch": 0.2506042488233049, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.157672882080078, "learning_rate": 8.346757100466298e-07, "loss": 0.4539, "mean_token_accuracy": 0.8530833721160889, "num_tokens": 75323912.0, "step": 1970 }, { "epoch": 0.2507314591018954, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.119637489318848, "learning_rate": 8.350996184824078e-07, "loss": 0.4944, "mean_token_accuracy": 0.8374307155609131, "num_tokens": 75361206.0, "step": 1971 }, { "epoch": 0.25085866938048595, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.157092094421387, "learning_rate": 8.355235269181856e-07, "loss": 0.475, "mean_token_accuracy": 0.8455842137336731, "num_tokens": 75397292.0, "step": 1972 }, { "epoch": 0.2509858796590764, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.105527877807617, "learning_rate": 8.359474353539635e-07, "loss": 0.4346, "mean_token_accuracy": 0.8589917421340942, "num_tokens": 75437616.0, "step": 1973 }, { "epoch": 0.25111308993766696, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.131040573120117, "learning_rate": 8.363713437897414e-07, "loss": 0.5245, "mean_token_accuracy": 0.8328005075454712, "num_tokens": 75486172.0, "step": 1974 }, { "epoch": 0.2512403002162575, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.1483154296875, "learning_rate": 8.367952522255193e-07, "loss": 0.4597, "mean_token_accuracy": 0.8477309942245483, "num_tokens": 75526637.0, "step": 1975 }, { "epoch": 0.25136751049484796, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.190511703491211, "learning_rate": 8.372191606612972e-07, "loss": 0.4614, "mean_token_accuracy": 0.8512825965881348, "num_tokens": 75564632.0, "step": 1976 }, { "epoch": 0.2514947207734385, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.194170951843262, "learning_rate": 8.376430690970749e-07, "loss": 0.4294, "mean_token_accuracy": 0.8613259196281433, "num_tokens": 75595553.0, "step": 1977 }, { "epoch": 0.251621931052029, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.142380714416504, "learning_rate": 8.380669775328528e-07, "loss": 0.4235, "mean_token_accuracy": 0.8631991147994995, "num_tokens": 75635847.0, "step": 1978 }, { "epoch": 0.25174914133061954, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.143804550170898, "learning_rate": 8.384908859686307e-07, "loss": 0.422, "mean_token_accuracy": 0.8642091751098633, "num_tokens": 75677403.0, "step": 1979 }, { "epoch": 0.25187635160921, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.147302627563477, "learning_rate": 8.389147944044086e-07, "loss": 0.4633, "mean_token_accuracy": 0.8489689826965332, "num_tokens": 75720703.0, "step": 1980 }, { "epoch": 0.25200356188780054, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.179489135742188, "learning_rate": 8.393387028401864e-07, "loss": 0.4873, "mean_token_accuracy": 0.8434510827064514, "num_tokens": 75758000.0, "step": 1981 }, { "epoch": 0.25213077216639107, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.137472152709961, "learning_rate": 8.397626112759644e-07, "loss": 0.5143, "mean_token_accuracy": 0.8347906470298767, "num_tokens": 75791113.0, "step": 1982 }, { "epoch": 0.25225798244498154, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.209735870361328, "learning_rate": 8.401865197117422e-07, "loss": 0.4596, "mean_token_accuracy": 0.8512554168701172, "num_tokens": 75821733.0, "step": 1983 }, { "epoch": 0.25238519272357207, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.154845237731934, "learning_rate": 8.406104281475202e-07, "loss": 0.4378, "mean_token_accuracy": 0.8555299043655396, "num_tokens": 75861404.0, "step": 1984 }, { "epoch": 0.2525124030021626, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.161663055419922, "learning_rate": 8.410343365832979e-07, "loss": 0.4389, "mean_token_accuracy": 0.8557260036468506, "num_tokens": 75903059.0, "step": 1985 }, { "epoch": 0.2526396132807531, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.141749382019043, "learning_rate": 8.414582450190758e-07, "loss": 0.4515, "mean_token_accuracy": 0.8535770177841187, "num_tokens": 75947125.0, "step": 1986 }, { "epoch": 0.2527668235593436, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.215758323669434, "learning_rate": 8.418821534548537e-07, "loss": 0.5356, "mean_token_accuracy": 0.8286312818527222, "num_tokens": 75987519.0, "step": 1987 }, { "epoch": 0.25289403383793413, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.185600280761719, "learning_rate": 8.423060618906316e-07, "loss": 0.4612, "mean_token_accuracy": 0.850506067276001, "num_tokens": 76027728.0, "step": 1988 }, { "epoch": 0.2530212441165246, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.157793998718262, "learning_rate": 8.427299703264095e-07, "loss": 0.4601, "mean_token_accuracy": 0.8520957231521606, "num_tokens": 76070942.0, "step": 1989 }, { "epoch": 0.25314845439511513, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.247318267822266, "learning_rate": 8.431538787621874e-07, "loss": 0.4859, "mean_token_accuracy": 0.8439293503761292, "num_tokens": 76113330.0, "step": 1990 }, { "epoch": 0.25327566467370566, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.15704345703125, "learning_rate": 8.435777871979652e-07, "loss": 0.439, "mean_token_accuracy": 0.8573590517044067, "num_tokens": 76153511.0, "step": 1991 }, { "epoch": 0.25340287495229613, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.266725540161133, "learning_rate": 8.440016956337432e-07, "loss": 0.4641, "mean_token_accuracy": 0.8507484197616577, "num_tokens": 76191505.0, "step": 1992 }, { "epoch": 0.25353008523088666, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.229636192321777, "learning_rate": 8.444256040695209e-07, "loss": 0.4802, "mean_token_accuracy": 0.8468472957611084, "num_tokens": 76232479.0, "step": 1993 }, { "epoch": 0.2536572955094772, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 8.187911033630371, "learning_rate": 8.448495125052988e-07, "loss": 0.4737, "mean_token_accuracy": 0.8481472730636597, "num_tokens": 76270192.0, "step": 1994 }, { "epoch": 0.25378450578806766, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.2694091796875, "learning_rate": 8.452734209410767e-07, "loss": 0.4142, "mean_token_accuracy": 0.8652669191360474, "num_tokens": 76303981.0, "step": 1995 }, { "epoch": 0.2539117160666582, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.185104370117188, "learning_rate": 8.456973293768545e-07, "loss": 0.4384, "mean_token_accuracy": 0.855819046497345, "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": 8.25745964050293, "learning_rate": 8.461212378126325e-07, "loss": 0.4924, "mean_token_accuracy": 0.8414432406425476, "num_tokens": 76377904.0, "step": 1997 }, { "epoch": 0.2541661366238392, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.240196228027344, "learning_rate": 8.465451462484103e-07, "loss": 0.5153, "mean_token_accuracy": 0.8363007307052612, "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": 8.23642635345459, "learning_rate": 8.469690546841882e-07, "loss": 0.4648, "mean_token_accuracy": 0.8487558364868164, "num_tokens": 76451591.0, "step": 1999 }, { "epoch": 0.25442055718102025, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.202631950378418, "learning_rate": 8.47392963119966e-07, "loss": 0.4446, "mean_token_accuracy": 0.8549876809120178, "num_tokens": 76486191.0, "step": 2000 }, { "epoch": 0.2545477674596107, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.232955932617188, "learning_rate": 8.478168715557439e-07, "loss": 0.4643, "mean_token_accuracy": 0.851065993309021, "num_tokens": 76523317.0, "step": 2001 }, { "epoch": 0.25467497773820125, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.219645500183105, "learning_rate": 8.482407799915217e-07, "loss": 0.471, "mean_token_accuracy": 0.8471846580505371, "num_tokens": 76562661.0, "step": 2002 }, { "epoch": 0.2548021880167918, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.215035438537598, "learning_rate": 8.486646884272997e-07, "loss": 0.4653, "mean_token_accuracy": 0.8468887209892273, "num_tokens": 76600914.0, "step": 2003 }, { "epoch": 0.25492939829538225, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.23685359954834, "learning_rate": 8.490885968630775e-07, "loss": 0.4851, "mean_token_accuracy": 0.8406269550323486, "num_tokens": 76636310.0, "step": 2004 }, { "epoch": 0.2550566085739728, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.248811721801758, "learning_rate": 8.495125052988555e-07, "loss": 0.4703, "mean_token_accuracy": 0.8502299785614014, "num_tokens": 76680889.0, "step": 2005 }, { "epoch": 0.2551838188525633, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.238907814025879, "learning_rate": 8.499364137346333e-07, "loss": 0.4625, "mean_token_accuracy": 0.8537341952323914, "num_tokens": 76722469.0, "step": 2006 }, { "epoch": 0.2553110291311538, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.255714416503906, "learning_rate": 8.503603221704112e-07, "loss": 0.4977, "mean_token_accuracy": 0.8390935063362122, "num_tokens": 76757668.0, "step": 2007 }, { "epoch": 0.2554382394097443, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.225812911987305, "learning_rate": 8.50784230606189e-07, "loss": 0.4586, "mean_token_accuracy": 0.8516169190406799, "num_tokens": 76795607.0, "step": 2008 }, { "epoch": 0.25556544968833483, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.218525886535645, "learning_rate": 8.512081390419669e-07, "loss": 0.4872, "mean_token_accuracy": 0.8497174978256226, "num_tokens": 76837161.0, "step": 2009 }, { "epoch": 0.2556926599669253, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.298903465270996, "learning_rate": 8.516320474777447e-07, "loss": 0.4668, "mean_token_accuracy": 0.8492447733879089, "num_tokens": 76872198.0, "step": 2010 }, { "epoch": 0.25581987024551583, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.278265953063965, "learning_rate": 8.520559559135227e-07, "loss": 0.4322, "mean_token_accuracy": 0.8622463345527649, "num_tokens": 76908976.0, "step": 2011 }, { "epoch": 0.25594708052410636, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.304243087768555, "learning_rate": 8.524798643493005e-07, "loss": 0.4831, "mean_token_accuracy": 0.8482089042663574, "num_tokens": 76947298.0, "step": 2012 }, { "epoch": 0.25607429080269684, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.26249885559082, "learning_rate": 8.529037727850785e-07, "loss": 0.4389, "mean_token_accuracy": 0.8551704287528992, "num_tokens": 76985112.0, "step": 2013 }, { "epoch": 0.25620150108128736, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.292221069335938, "learning_rate": 8.533276812208563e-07, "loss": 0.4119, "mean_token_accuracy": 0.864578127861023, "num_tokens": 77020309.0, "step": 2014 }, { "epoch": 0.2563287113598779, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.265891075134277, "learning_rate": 8.53751589656634e-07, "loss": 0.4516, "mean_token_accuracy": 0.850540041923523, "num_tokens": 77063245.0, "step": 2015 }, { "epoch": 0.25645592163846836, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 8.300527572631836, "learning_rate": 8.54175498092412e-07, "loss": 0.449, "mean_token_accuracy": 0.853595495223999, "num_tokens": 77096791.0, "step": 2016 }, { "epoch": 0.2565831319170589, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.33018684387207, "learning_rate": 8.545994065281898e-07, "loss": 0.4282, "mean_token_accuracy": 0.8615337014198303, "num_tokens": 77129281.0, "step": 2017 }, { "epoch": 0.2567103421956494, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.261917114257812, "learning_rate": 8.550233149639677e-07, "loss": 0.5156, "mean_token_accuracy": 0.8316926956176758, "num_tokens": 77168785.0, "step": 2018 }, { "epoch": 0.2568375524742399, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.288542747497559, "learning_rate": 8.554472233997456e-07, "loss": 0.4833, "mean_token_accuracy": 0.8408437967300415, "num_tokens": 77208360.0, "step": 2019 }, { "epoch": 0.2569647627528304, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.26332950592041, "learning_rate": 8.558711318355235e-07, "loss": 0.5158, "mean_token_accuracy": 0.8323491215705872, "num_tokens": 77245539.0, "step": 2020 }, { "epoch": 0.25709197303142095, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.293261528015137, "learning_rate": 8.562950402713014e-07, "loss": 0.4188, "mean_token_accuracy": 0.8682722449302673, "num_tokens": 77279118.0, "step": 2021 }, { "epoch": 0.2572191833100114, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.278204917907715, "learning_rate": 8.567189487070793e-07, "loss": 0.5059, "mean_token_accuracy": 0.8400536775588989, "num_tokens": 77316259.0, "step": 2022 }, { "epoch": 0.25734639358860195, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.249100685119629, "learning_rate": 8.57142857142857e-07, "loss": 0.5228, "mean_token_accuracy": 0.8322848081588745, "num_tokens": 77359712.0, "step": 2023 }, { "epoch": 0.2574736038671925, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.260873794555664, "learning_rate": 8.57566765578635e-07, "loss": 0.4271, "mean_token_accuracy": 0.8641562461853027, "num_tokens": 77402269.0, "step": 2024 }, { "epoch": 0.25760081414578295, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.246637344360352, "learning_rate": 8.579906740144128e-07, "loss": 0.4555, "mean_token_accuracy": 0.8530021905899048, "num_tokens": 77445791.0, "step": 2025 }, { "epoch": 0.2577280244243735, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.290799140930176, "learning_rate": 8.584145824501907e-07, "loss": 0.489, "mean_token_accuracy": 0.8395321369171143, "num_tokens": 77485376.0, "step": 2026 }, { "epoch": 0.257855234702964, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.279889106750488, "learning_rate": 8.588384908859686e-07, "loss": 0.4851, "mean_token_accuracy": 0.8459616899490356, "num_tokens": 77524344.0, "step": 2027 }, { "epoch": 0.25798244498155454, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 8.272439956665039, "learning_rate": 8.592623993217465e-07, "loss": 0.4788, "mean_token_accuracy": 0.8472429513931274, "num_tokens": 77562746.0, "step": 2028 }, { "epoch": 0.258109655260145, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.28226089477539, "learning_rate": 8.596863077575244e-07, "loss": 0.501, "mean_token_accuracy": 0.8418809175491333, "num_tokens": 77601261.0, "step": 2029 }, { "epoch": 0.25823686553873554, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.297844886779785, "learning_rate": 8.601102161933023e-07, "loss": 0.4818, "mean_token_accuracy": 0.844495415687561, "num_tokens": 77640054.0, "step": 2030 }, { "epoch": 0.25836407581732607, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.288308143615723, "learning_rate": 8.6053412462908e-07, "loss": 0.4561, "mean_token_accuracy": 0.8526021242141724, "num_tokens": 77680098.0, "step": 2031 }, { "epoch": 0.25849128609591654, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.340460777282715, "learning_rate": 8.60958033064858e-07, "loss": 0.4501, "mean_token_accuracy": 0.8557475805282593, "num_tokens": 77712517.0, "step": 2032 }, { "epoch": 0.25861849637450707, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.31368350982666, "learning_rate": 8.613819415006358e-07, "loss": 0.512, "mean_token_accuracy": 0.8414268493652344, "num_tokens": 77751335.0, "step": 2033 }, { "epoch": 0.2587457066530976, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.271329879760742, "learning_rate": 8.618058499364137e-07, "loss": 0.4994, "mean_token_accuracy": 0.8391794562339783, "num_tokens": 77792300.0, "step": 2034 }, { "epoch": 0.25887291693168807, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.300090789794922, "learning_rate": 8.622297583721916e-07, "loss": 0.4579, "mean_token_accuracy": 0.8550108671188354, "num_tokens": 77836391.0, "step": 2035 }, { "epoch": 0.2590001272102786, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.302776336669922, "learning_rate": 8.626536668079695e-07, "loss": 0.4686, "mean_token_accuracy": 0.84682297706604, "num_tokens": 77871600.0, "step": 2036 }, { "epoch": 0.2591273374888691, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.355738639831543, "learning_rate": 8.630775752437474e-07, "loss": 0.4567, "mean_token_accuracy": 0.8528156876564026, "num_tokens": 77904623.0, "step": 2037 }, { "epoch": 0.2592545477674596, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.299013137817383, "learning_rate": 8.635014836795251e-07, "loss": 0.4956, "mean_token_accuracy": 0.8414185047149658, "num_tokens": 77941145.0, "step": 2038 }, { "epoch": 0.2593817580460501, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.355462074279785, "learning_rate": 8.63925392115303e-07, "loss": 0.4334, "mean_token_accuracy": 0.8568742871284485, "num_tokens": 77975087.0, "step": 2039 }, { "epoch": 0.25950896832464065, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.26848030090332, "learning_rate": 8.643493005510809e-07, "loss": 0.4405, "mean_token_accuracy": 0.8560633659362793, "num_tokens": 78009526.0, "step": 2040 }, { "epoch": 0.2596361786032311, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.33358097076416, "learning_rate": 8.647732089868588e-07, "loss": 0.5183, "mean_token_accuracy": 0.8321500420570374, "num_tokens": 78050918.0, "step": 2041 }, { "epoch": 0.25976338888182166, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.324875831604004, "learning_rate": 8.651971174226366e-07, "loss": 0.4353, "mean_token_accuracy": 0.8586117029190063, "num_tokens": 78085547.0, "step": 2042 }, { "epoch": 0.2598905991604122, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 8.373218536376953, "learning_rate": 8.656210258584146e-07, "loss": 0.5041, "mean_token_accuracy": 0.8361818790435791, "num_tokens": 78122916.0, "step": 2043 }, { "epoch": 0.26001780943900266, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.332290649414062, "learning_rate": 8.660449342941924e-07, "loss": 0.4968, "mean_token_accuracy": 0.8367646336555481, "num_tokens": 78161800.0, "step": 2044 }, { "epoch": 0.2601450197175932, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 8.325106620788574, "learning_rate": 8.664688427299704e-07, "loss": 0.4755, "mean_token_accuracy": 0.846686601638794, "num_tokens": 78199961.0, "step": 2045 }, { "epoch": 0.2602722299961837, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.372488975524902, "learning_rate": 8.668927511657481e-07, "loss": 0.4331, "mean_token_accuracy": 0.8569824695587158, "num_tokens": 78231332.0, "step": 2046 }, { "epoch": 0.2603994402747742, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 8.367506980895996, "learning_rate": 8.67316659601526e-07, "loss": 0.4895, "mean_token_accuracy": 0.8409151434898376, "num_tokens": 78265054.0, "step": 2047 }, { "epoch": 0.2605266505533647, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.303439140319824, "learning_rate": 8.677405680373039e-07, "loss": 0.4461, "mean_token_accuracy": 0.8575054407119751, "num_tokens": 78305445.0, "step": 2048 }, { "epoch": 0.26065386083195524, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.381278038024902, "learning_rate": 8.681644764730818e-07, "loss": 0.4754, "mean_token_accuracy": 0.8457299470901489, "num_tokens": 78339174.0, "step": 2049 }, { "epoch": 0.2607810711105457, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.333961486816406, "learning_rate": 8.685883849088596e-07, "loss": 0.4182, "mean_token_accuracy": 0.8635318279266357, "num_tokens": 78377630.0, "step": 2050 }, { "epoch": 0.26090828138913624, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.300654411315918, "learning_rate": 8.690122933446376e-07, "loss": 0.475, "mean_token_accuracy": 0.8491186499595642, "num_tokens": 78419908.0, "step": 2051 }, { "epoch": 0.26103549166772677, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.341570854187012, "learning_rate": 8.694362017804154e-07, "loss": 0.4412, "mean_token_accuracy": 0.8583394289016724, "num_tokens": 78459561.0, "step": 2052 }, { "epoch": 0.26116270194631724, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.354602813720703, "learning_rate": 8.698601102161933e-07, "loss": 0.4729, "mean_token_accuracy": 0.8467732071876526, "num_tokens": 78502500.0, "step": 2053 }, { "epoch": 0.26128991222490777, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.357635498046875, "learning_rate": 8.702840186519711e-07, "loss": 0.4306, "mean_token_accuracy": 0.8602506518363953, "num_tokens": 78545411.0, "step": 2054 }, { "epoch": 0.2614171225034983, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.369023323059082, "learning_rate": 8.70707927087749e-07, "loss": 0.4299, "mean_token_accuracy": 0.860849916934967, "num_tokens": 78585878.0, "step": 2055 }, { "epoch": 0.2615443327820888, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.366920471191406, "learning_rate": 8.711318355235269e-07, "loss": 0.5188, "mean_token_accuracy": 0.8301746845245361, "num_tokens": 78621233.0, "step": 2056 }, { "epoch": 0.2616715430606793, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.411080360412598, "learning_rate": 8.715557439593047e-07, "loss": 0.4391, "mean_token_accuracy": 0.8553622961044312, "num_tokens": 78654644.0, "step": 2057 }, { "epoch": 0.26179875333926983, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.384950637817383, "learning_rate": 8.719796523950826e-07, "loss": 0.5157, "mean_token_accuracy": 0.8322256803512573, "num_tokens": 78692851.0, "step": 2058 }, { "epoch": 0.2619259636178603, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.387900352478027, "learning_rate": 8.724035608308605e-07, "loss": 0.4093, "mean_token_accuracy": 0.8662121295928955, "num_tokens": 78725329.0, "step": 2059 }, { "epoch": 0.26205317389645083, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.367546081542969, "learning_rate": 8.728274692666384e-07, "loss": 0.4661, "mean_token_accuracy": 0.8500974774360657, "num_tokens": 78766645.0, "step": 2060 }, { "epoch": 0.26218038417504136, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.3793306350708, "learning_rate": 8.732513777024162e-07, "loss": 0.4609, "mean_token_accuracy": 0.851863443851471, "num_tokens": 78801576.0, "step": 2061 }, { "epoch": 0.26230759445363183, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.346324920654297, "learning_rate": 8.736752861381941e-07, "loss": 0.5039, "mean_token_accuracy": 0.8371524214744568, "num_tokens": 78845577.0, "step": 2062 }, { "epoch": 0.26243480473222236, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.426884651184082, "learning_rate": 8.740991945739719e-07, "loss": 0.4693, "mean_token_accuracy": 0.8486618995666504, "num_tokens": 78884229.0, "step": 2063 }, { "epoch": 0.2625620150108129, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.4127779006958, "learning_rate": 8.745231030097499e-07, "loss": 0.4562, "mean_token_accuracy": 0.8527231216430664, "num_tokens": 78926948.0, "step": 2064 }, { "epoch": 0.26268922528940336, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.428046226501465, "learning_rate": 8.749470114455277e-07, "loss": 0.4324, "mean_token_accuracy": 0.8583706617355347, "num_tokens": 78963582.0, "step": 2065 }, { "epoch": 0.2628164355679939, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.394137382507324, "learning_rate": 8.753709198813056e-07, "loss": 0.5394, "mean_token_accuracy": 0.8266852498054504, "num_tokens": 79007345.0, "step": 2066 }, { "epoch": 0.2629436458465844, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.40161418914795, "learning_rate": 8.757948283170835e-07, "loss": 0.4376, "mean_token_accuracy": 0.8599448204040527, "num_tokens": 79049505.0, "step": 2067 }, { "epoch": 0.2630708561251749, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.440741539001465, "learning_rate": 8.762187367528613e-07, "loss": 0.432, "mean_token_accuracy": 0.8561379313468933, "num_tokens": 79087419.0, "step": 2068 }, { "epoch": 0.2631980664037654, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.402225494384766, "learning_rate": 8.766426451886392e-07, "loss": 0.4799, "mean_token_accuracy": 0.8430930376052856, "num_tokens": 79131444.0, "step": 2069 }, { "epoch": 0.26332527668235595, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.39874267578125, "learning_rate": 8.770665536244171e-07, "loss": 0.5064, "mean_token_accuracy": 0.8396645784378052, "num_tokens": 79171976.0, "step": 2070 }, { "epoch": 0.2634524869609464, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.43625545501709, "learning_rate": 8.774904620601949e-07, "loss": 0.4353, "mean_token_accuracy": 0.8563985824584961, "num_tokens": 79206273.0, "step": 2071 }, { "epoch": 0.26357969723953695, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.36558723449707, "learning_rate": 8.779143704959729e-07, "loss": 0.4359, "mean_token_accuracy": 0.8576169013977051, "num_tokens": 79245564.0, "step": 2072 }, { "epoch": 0.2637069075181275, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 8.438196182250977, "learning_rate": 8.783382789317507e-07, "loss": 0.4675, "mean_token_accuracy": 0.8493080139160156, "num_tokens": 79281857.0, "step": 2073 }, { "epoch": 0.26383411779671795, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.396097183227539, "learning_rate": 8.787621873675286e-07, "loss": 0.5009, "mean_token_accuracy": 0.8345674872398376, "num_tokens": 79312482.0, "step": 2074 }, { "epoch": 0.2639613280753085, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 8.439184188842773, "learning_rate": 8.791860958033065e-07, "loss": 0.4401, "mean_token_accuracy": 0.8574779033660889, "num_tokens": 79348157.0, "step": 2075 }, { "epoch": 0.264088538353899, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.348613739013672, "learning_rate": 8.796100042390842e-07, "loss": 0.4266, "mean_token_accuracy": 0.8615070581436157, "num_tokens": 79387540.0, "step": 2076 }, { "epoch": 0.2642157486324895, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 8.468825340270996, "learning_rate": 8.800339126748622e-07, "loss": 0.4435, "mean_token_accuracy": 0.8543331027030945, "num_tokens": 79421424.0, "step": 2077 }, { "epoch": 0.26434295891108, "ewc_loss": 0.010498046875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 8.392165184020996, "learning_rate": 8.8045782111064e-07, "loss": 0.5143, "mean_token_accuracy": 0.8335309624671936, "num_tokens": 79458733.0, "step": 2078 }, { "epoch": 0.26447016918967053, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.386728286743164, "learning_rate": 8.808817295464179e-07, "loss": 0.4661, "mean_token_accuracy": 0.850365400314331, "num_tokens": 79495394.0, "step": 2079 }, { "epoch": 0.26459737946826106, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.420526504516602, "learning_rate": 8.813056379821958e-07, "loss": 0.4374, "mean_token_accuracy": 0.8592714071273804, "num_tokens": 79529799.0, "step": 2080 }, { "epoch": 0.26472458974685154, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.4119873046875, "learning_rate": 8.817295464179737e-07, "loss": 0.4629, "mean_token_accuracy": 0.8500411510467529, "num_tokens": 79564664.0, "step": 2081 }, { "epoch": 0.26485180002544206, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.46022891998291, "learning_rate": 8.821534548537515e-07, "loss": 0.5405, "mean_token_accuracy": 0.8239262700080872, "num_tokens": 79606104.0, "step": 2082 }, { "epoch": 0.2649790103040326, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.390202522277832, "learning_rate": 8.825773632895295e-07, "loss": 0.4366, "mean_token_accuracy": 0.858872652053833, "num_tokens": 79647688.0, "step": 2083 }, { "epoch": 0.26510622058262306, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.417728424072266, "learning_rate": 8.830012717253072e-07, "loss": 0.4539, "mean_token_accuracy": 0.8553991317749023, "num_tokens": 79685500.0, "step": 2084 }, { "epoch": 0.2652334308612136, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.399410247802734, "learning_rate": 8.834251801610852e-07, "loss": 0.4825, "mean_token_accuracy": 0.843569278717041, "num_tokens": 79729912.0, "step": 2085 }, { "epoch": 0.2653606411398041, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.424172401428223, "learning_rate": 8.83849088596863e-07, "loss": 0.4384, "mean_token_accuracy": 0.8581832051277161, "num_tokens": 79767487.0, "step": 2086 }, { "epoch": 0.2654878514183946, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.476139068603516, "learning_rate": 8.842729970326409e-07, "loss": 0.4192, "mean_token_accuracy": 0.8659485578536987, "num_tokens": 79804802.0, "step": 2087 }, { "epoch": 0.2656150616969851, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.434623718261719, "learning_rate": 8.846969054684188e-07, "loss": 0.5368, "mean_token_accuracy": 0.8295632600784302, "num_tokens": 79845685.0, "step": 2088 }, { "epoch": 0.26574227197557565, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.426644325256348, "learning_rate": 8.851208139041967e-07, "loss": 0.4256, "mean_token_accuracy": 0.8606764078140259, "num_tokens": 79884692.0, "step": 2089 }, { "epoch": 0.2658694822541661, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 8.502670288085938, "learning_rate": 8.855447223399745e-07, "loss": 0.4851, "mean_token_accuracy": 0.8411546945571899, "num_tokens": 79919780.0, "step": 2090 }, { "epoch": 0.26599669253275665, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.45715045928955, "learning_rate": 8.859686307757524e-07, "loss": 0.511, "mean_token_accuracy": 0.8350951075553894, "num_tokens": 79958317.0, "step": 2091 }, { "epoch": 0.2661239028113472, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.42744255065918, "learning_rate": 8.863925392115302e-07, "loss": 0.4439, "mean_token_accuracy": 0.8546361923217773, "num_tokens": 79994126.0, "step": 2092 }, { "epoch": 0.26625111308993765, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.406499862670898, "learning_rate": 8.868164476473082e-07, "loss": 0.4136, "mean_token_accuracy": 0.8655385971069336, "num_tokens": 80033157.0, "step": 2093 }, { "epoch": 0.2663783233685282, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.412325859069824, "learning_rate": 8.87240356083086e-07, "loss": 0.4426, "mean_token_accuracy": 0.8566440939903259, "num_tokens": 80073354.0, "step": 2094 }, { "epoch": 0.2665055336471187, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.494939804077148, "learning_rate": 8.876642645188639e-07, "loss": 0.5032, "mean_token_accuracy": 0.837437629699707, "num_tokens": 80111819.0, "step": 2095 }, { "epoch": 0.2666327439257092, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.452038764953613, "learning_rate": 8.880881729546418e-07, "loss": 0.4174, "mean_token_accuracy": 0.8635067343711853, "num_tokens": 80147951.0, "step": 2096 }, { "epoch": 0.2667599542042997, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.416539192199707, "learning_rate": 8.885120813904197e-07, "loss": 0.4322, "mean_token_accuracy": 0.8582898378372192, "num_tokens": 80191319.0, "step": 2097 }, { "epoch": 0.26688716448289024, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.549505233764648, "learning_rate": 8.889359898261976e-07, "loss": 0.5189, "mean_token_accuracy": 0.8370068669319153, "num_tokens": 80226300.0, "step": 2098 }, { "epoch": 0.2670143747614807, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.441547393798828, "learning_rate": 8.893598982619753e-07, "loss": 0.4441, "mean_token_accuracy": 0.8551505208015442, "num_tokens": 80265465.0, "step": 2099 }, { "epoch": 0.26714158504007124, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.424515724182129, "learning_rate": 8.897838066977532e-07, "loss": 0.4471, "mean_token_accuracy": 0.8551241755485535, "num_tokens": 80312696.0, "step": 2100 }, { "epoch": 0.26726879531866177, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.484600067138672, "learning_rate": 8.902077151335311e-07, "loss": 0.4671, "mean_token_accuracy": 0.8468714952468872, "num_tokens": 80353507.0, "step": 2101 }, { "epoch": 0.26739600559725224, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.477831840515137, "learning_rate": 8.90631623569309e-07, "loss": 0.5503, "mean_token_accuracy": 0.8198553323745728, "num_tokens": 80390795.0, "step": 2102 }, { "epoch": 0.26752321587584277, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.490058898925781, "learning_rate": 8.910555320050868e-07, "loss": 0.4596, "mean_token_accuracy": 0.8511356115341187, "num_tokens": 80427599.0, "step": 2103 }, { "epoch": 0.2676504261544333, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.499859809875488, "learning_rate": 8.914794404408648e-07, "loss": 0.4624, "mean_token_accuracy": 0.8507114052772522, "num_tokens": 80468084.0, "step": 2104 }, { "epoch": 0.26777763643302377, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.40566349029541, "learning_rate": 8.919033488766426e-07, "loss": 0.4061, "mean_token_accuracy": 0.8683033585548401, "num_tokens": 80507218.0, "step": 2105 }, { "epoch": 0.2679048467116143, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.552169799804688, "learning_rate": 8.923272573124204e-07, "loss": 0.4901, "mean_token_accuracy": 0.8431082367897034, "num_tokens": 80541730.0, "step": 2106 }, { "epoch": 0.2680320569902048, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.457978248596191, "learning_rate": 8.927511657481983e-07, "loss": 0.501, "mean_token_accuracy": 0.8379601240158081, "num_tokens": 80584677.0, "step": 2107 }, { "epoch": 0.2681592672687953, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.474754333496094, "learning_rate": 8.931750741839762e-07, "loss": 0.4358, "mean_token_accuracy": 0.8607635498046875, "num_tokens": 80616332.0, "step": 2108 }, { "epoch": 0.2682864775473858, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.490438461303711, "learning_rate": 8.935989826197541e-07, "loss": 0.4445, "mean_token_accuracy": 0.8579303622245789, "num_tokens": 80654435.0, "step": 2109 }, { "epoch": 0.26841368782597635, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.45287799835205, "learning_rate": 8.94022891055532e-07, "loss": 0.4212, "mean_token_accuracy": 0.8665674328804016, "num_tokens": 80691246.0, "step": 2110 }, { "epoch": 0.2685408981045668, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.459409713745117, "learning_rate": 8.944467994913098e-07, "loss": 0.4722, "mean_token_accuracy": 0.8435226678848267, "num_tokens": 80730510.0, "step": 2111 }, { "epoch": 0.26866810838315736, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.430826187133789, "learning_rate": 8.948707079270878e-07, "loss": 0.3909, "mean_token_accuracy": 0.8716886043548584, "num_tokens": 80769369.0, "step": 2112 }, { "epoch": 0.2687953186617479, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.53916072845459, "learning_rate": 8.952946163628656e-07, "loss": 0.4456, "mean_token_accuracy": 0.8547161817550659, "num_tokens": 80802168.0, "step": 2113 }, { "epoch": 0.26892252894033836, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.5078125, "learning_rate": 8.957185247986434e-07, "loss": 0.396, "mean_token_accuracy": 0.8711913824081421, "num_tokens": 80839389.0, "step": 2114 }, { "epoch": 0.2690497392189289, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.50944709777832, "learning_rate": 8.961424332344213e-07, "loss": 0.4511, "mean_token_accuracy": 0.8550161123275757, "num_tokens": 80881139.0, "step": 2115 }, { "epoch": 0.2691769494975194, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.508222579956055, "learning_rate": 8.965663416701992e-07, "loss": 0.4527, "mean_token_accuracy": 0.8514912128448486, "num_tokens": 80919809.0, "step": 2116 }, { "epoch": 0.2693041597761099, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 8.593765258789062, "learning_rate": 8.969902501059771e-07, "loss": 0.4289, "mean_token_accuracy": 0.863511323928833, "num_tokens": 80956860.0, "step": 2117 }, { "epoch": 0.2694313700547004, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.50696849822998, "learning_rate": 8.97414158541755e-07, "loss": 0.3958, "mean_token_accuracy": 0.8706482648849487, "num_tokens": 80994322.0, "step": 2118 }, { "epoch": 0.26955858033329094, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.585421562194824, "learning_rate": 8.978380669775328e-07, "loss": 0.479, "mean_token_accuracy": 0.844490647315979, "num_tokens": 81032792.0, "step": 2119 }, { "epoch": 0.2696857906118814, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.533742904663086, "learning_rate": 8.982619754133107e-07, "loss": 0.4789, "mean_token_accuracy": 0.8463651537895203, "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": 8.494179725646973, "learning_rate": 8.986858838490886e-07, "loss": 0.4813, "mean_token_accuracy": 0.8453303575515747, "num_tokens": 81113145.0, "step": 2121 }, { "epoch": 0.26994021116906247, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.557648658752441, "learning_rate": 8.991097922848663e-07, "loss": 0.4499, "mean_token_accuracy": 0.8554574251174927, "num_tokens": 81154175.0, "step": 2122 }, { "epoch": 0.27006742144765294, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.496938705444336, "learning_rate": 8.995337007206443e-07, "loss": 0.4739, "mean_token_accuracy": 0.8485828042030334, "num_tokens": 81193319.0, "step": 2123 }, { "epoch": 0.2701946317262435, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.550846099853516, "learning_rate": 8.999576091564221e-07, "loss": 0.4412, "mean_token_accuracy": 0.8574445247650146, "num_tokens": 81229057.0, "step": 2124 }, { "epoch": 0.270321842004834, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.54643726348877, "learning_rate": 9.003815175922001e-07, "loss": 0.4495, "mean_token_accuracy": 0.8538724184036255, "num_tokens": 81262380.0, "step": 2125 }, { "epoch": 0.2704490522834245, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.55631160736084, "learning_rate": 9.008054260279779e-07, "loss": 0.4542, "mean_token_accuracy": 0.8540869951248169, "num_tokens": 81304080.0, "step": 2126 }, { "epoch": 0.270576262562015, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.581528663635254, "learning_rate": 9.012293344637558e-07, "loss": 0.4841, "mean_token_accuracy": 0.8455721139907837, "num_tokens": 81340226.0, "step": 2127 }, { "epoch": 0.27070347284060553, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 8.526460647583008, "learning_rate": 9.016532428995337e-07, "loss": 0.4308, "mean_token_accuracy": 0.8584364056587219, "num_tokens": 81378046.0, "step": 2128 }, { "epoch": 0.27083068311919606, "ewc_loss": 0.0107421875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 8.535730361938477, "learning_rate": 9.020771513353115e-07, "loss": 0.4751, "mean_token_accuracy": 0.8469637036323547, "num_tokens": 81419486.0, "step": 2129 }, { "epoch": 0.27095789339778653, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.554191589355469, "learning_rate": 9.025010597710894e-07, "loss": 0.4742, "mean_token_accuracy": 0.8465868234634399, "num_tokens": 81457488.0, "step": 2130 }, { "epoch": 0.27108510367637706, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.539231300354004, "learning_rate": 9.029249682068673e-07, "loss": 0.4428, "mean_token_accuracy": 0.8565492630004883, "num_tokens": 81494812.0, "step": 2131 }, { "epoch": 0.2712123139549676, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.514432907104492, "learning_rate": 9.033488766426451e-07, "loss": 0.4596, "mean_token_accuracy": 0.8499321341514587, "num_tokens": 81531421.0, "step": 2132 }, { "epoch": 0.27133952423355806, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.564778327941895, "learning_rate": 9.037727850784231e-07, "loss": 0.4252, "mean_token_accuracy": 0.8603631854057312, "num_tokens": 81567221.0, "step": 2133 }, { "epoch": 0.2714667345121486, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.558053016662598, "learning_rate": 9.041966935142009e-07, "loss": 0.4599, "mean_token_accuracy": 0.8517550230026245, "num_tokens": 81605832.0, "step": 2134 }, { "epoch": 0.2715939447907391, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.532827377319336, "learning_rate": 9.046206019499788e-07, "loss": 0.4151, "mean_token_accuracy": 0.8670586347579956, "num_tokens": 81646919.0, "step": 2135 }, { "epoch": 0.2717211550693296, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.567989349365234, "learning_rate": 9.050445103857567e-07, "loss": 0.4926, "mean_token_accuracy": 0.8455215692520142, "num_tokens": 81687214.0, "step": 2136 }, { "epoch": 0.2718483653479201, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.599026679992676, "learning_rate": 9.054684188215344e-07, "loss": 0.3981, "mean_token_accuracy": 0.8709843754768372, "num_tokens": 81721437.0, "step": 2137 }, { "epoch": 0.27197557562651065, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.613603591918945, "learning_rate": 9.058923272573124e-07, "loss": 0.4723, "mean_token_accuracy": 0.8472138047218323, "num_tokens": 81756595.0, "step": 2138 }, { "epoch": 0.2721027859051011, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.537360191345215, "learning_rate": 9.063162356930902e-07, "loss": 0.4362, "mean_token_accuracy": 0.8599505424499512, "num_tokens": 81796198.0, "step": 2139 }, { "epoch": 0.27222999618369165, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.54450511932373, "learning_rate": 9.067401441288681e-07, "loss": 0.4677, "mean_token_accuracy": 0.851409912109375, "num_tokens": 81833002.0, "step": 2140 }, { "epoch": 0.2723572064622822, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.580757141113281, "learning_rate": 9.07164052564646e-07, "loss": 0.4393, "mean_token_accuracy": 0.8531270027160645, "num_tokens": 81872089.0, "step": 2141 }, { "epoch": 0.27248441674087265, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.613466262817383, "learning_rate": 9.075879610004239e-07, "loss": 0.475, "mean_token_accuracy": 0.8456190824508667, "num_tokens": 81908356.0, "step": 2142 }, { "epoch": 0.2726116270194632, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.574603080749512, "learning_rate": 9.080118694362017e-07, "loss": 0.4793, "mean_token_accuracy": 0.8433582782745361, "num_tokens": 81950251.0, "step": 2143 }, { "epoch": 0.2727388372980537, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.629571914672852, "learning_rate": 9.084357778719796e-07, "loss": 0.472, "mean_token_accuracy": 0.8450815677642822, "num_tokens": 81987157.0, "step": 2144 }, { "epoch": 0.2728660475766442, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.604663848876953, "learning_rate": 9.088596863077574e-07, "loss": 0.4565, "mean_token_accuracy": 0.8554622530937195, "num_tokens": 82023633.0, "step": 2145 }, { "epoch": 0.2729932578552347, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.59477424621582, "learning_rate": 9.092835947435354e-07, "loss": 0.4585, "mean_token_accuracy": 0.8476675748825073, "num_tokens": 82061311.0, "step": 2146 }, { "epoch": 0.27312046813382523, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.608196258544922, "learning_rate": 9.097075031793132e-07, "loss": 0.4934, "mean_token_accuracy": 0.8417124152183533, "num_tokens": 82098370.0, "step": 2147 }, { "epoch": 0.2732476784124157, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 8.56607437133789, "learning_rate": 9.101314116150911e-07, "loss": 0.4213, "mean_token_accuracy": 0.867032527923584, "num_tokens": 82134931.0, "step": 2148 }, { "epoch": 0.27337488869100623, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.581700325012207, "learning_rate": 9.10555320050869e-07, "loss": 0.4301, "mean_token_accuracy": 0.8607032299041748, "num_tokens": 82171391.0, "step": 2149 }, { "epoch": 0.27350209896959676, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.541987419128418, "learning_rate": 9.109792284866469e-07, "loss": 0.4495, "mean_token_accuracy": 0.8584659099578857, "num_tokens": 82214064.0, "step": 2150 }, { "epoch": 0.27362930924818724, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.592241287231445, "learning_rate": 9.114031369224247e-07, "loss": 0.4648, "mean_token_accuracy": 0.8497268557548523, "num_tokens": 82252774.0, "step": 2151 }, { "epoch": 0.27375651952677776, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.617551803588867, "learning_rate": 9.118270453582026e-07, "loss": 0.5203, "mean_token_accuracy": 0.8368427753448486, "num_tokens": 82291645.0, "step": 2152 }, { "epoch": 0.2738837298053683, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.657288551330566, "learning_rate": 9.122509537939804e-07, "loss": 0.4526, "mean_token_accuracy": 0.8479210734367371, "num_tokens": 82333643.0, "step": 2153 }, { "epoch": 0.27401094008395877, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.569662094116211, "learning_rate": 9.126748622297584e-07, "loss": 0.4906, "mean_token_accuracy": 0.8383809328079224, "num_tokens": 82370792.0, "step": 2154 }, { "epoch": 0.2741381503625493, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.615796089172363, "learning_rate": 9.130987706655362e-07, "loss": 0.4798, "mean_token_accuracy": 0.8442689180374146, "num_tokens": 82408626.0, "step": 2155 }, { "epoch": 0.2742653606411398, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.517772674560547, "learning_rate": 9.135226791013141e-07, "loss": 0.5002, "mean_token_accuracy": 0.8443499803543091, "num_tokens": 82448401.0, "step": 2156 }, { "epoch": 0.2743925709197303, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.64752197265625, "learning_rate": 9.13946587537092e-07, "loss": 0.4569, "mean_token_accuracy": 0.8480283617973328, "num_tokens": 82476644.0, "step": 2157 }, { "epoch": 0.2745197811983208, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.59597396850586, "learning_rate": 9.143704959728699e-07, "loss": 0.4726, "mean_token_accuracy": 0.8430241942405701, "num_tokens": 82512722.0, "step": 2158 }, { "epoch": 0.27464699147691135, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.594927787780762, "learning_rate": 9.147944044086476e-07, "loss": 0.4922, "mean_token_accuracy": 0.8413487672805786, "num_tokens": 82545393.0, "step": 2159 }, { "epoch": 0.2747742017555018, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.593526840209961, "learning_rate": 9.152183128444255e-07, "loss": 0.4894, "mean_token_accuracy": 0.8476650714874268, "num_tokens": 82586329.0, "step": 2160 }, { "epoch": 0.27490141203409235, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.593125343322754, "learning_rate": 9.156422212802034e-07, "loss": 0.4964, "mean_token_accuracy": 0.8366557955741882, "num_tokens": 82625563.0, "step": 2161 }, { "epoch": 0.2750286223126829, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.652257919311523, "learning_rate": 9.160661297159813e-07, "loss": 0.4346, "mean_token_accuracy": 0.8589305877685547, "num_tokens": 82660030.0, "step": 2162 }, { "epoch": 0.27515583259127335, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.631937980651855, "learning_rate": 9.164900381517592e-07, "loss": 0.4566, "mean_token_accuracy": 0.8515449166297913, "num_tokens": 82698794.0, "step": 2163 }, { "epoch": 0.2752830428698639, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.646980285644531, "learning_rate": 9.16913946587537e-07, "loss": 0.455, "mean_token_accuracy": 0.8501067757606506, "num_tokens": 82740015.0, "step": 2164 }, { "epoch": 0.2754102531484544, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.689109802246094, "learning_rate": 9.17337855023315e-07, "loss": 0.4685, "mean_token_accuracy": 0.8476443886756897, "num_tokens": 82775765.0, "step": 2165 }, { "epoch": 0.2755374634270449, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.628390312194824, "learning_rate": 9.177617634590928e-07, "loss": 0.4196, "mean_token_accuracy": 0.8620436787605286, "num_tokens": 82814324.0, "step": 2166 }, { "epoch": 0.2756646737056354, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.613033294677734, "learning_rate": 9.181856718948706e-07, "loss": 0.4068, "mean_token_accuracy": 0.8654816150665283, "num_tokens": 82851007.0, "step": 2167 }, { "epoch": 0.27579188398422594, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.701961517333984, "learning_rate": 9.186095803306485e-07, "loss": 0.4501, "mean_token_accuracy": 0.8532503843307495, "num_tokens": 82886846.0, "step": 2168 }, { "epoch": 0.2759190942628164, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 8.650602340698242, "learning_rate": 9.190334887664264e-07, "loss": 0.488, "mean_token_accuracy": 0.842734158039093, "num_tokens": 82929607.0, "step": 2169 }, { "epoch": 0.27604630454140694, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.720453262329102, "learning_rate": 9.194573972022043e-07, "loss": 0.5454, "mean_token_accuracy": 0.8315541744232178, "num_tokens": 82965089.0, "step": 2170 }, { "epoch": 0.27617351481999747, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.618912696838379, "learning_rate": 9.198813056379822e-07, "loss": 0.5201, "mean_token_accuracy": 0.8318504691123962, "num_tokens": 83008345.0, "step": 2171 }, { "epoch": 0.27630072509858794, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.706275939941406, "learning_rate": 9.2030521407376e-07, "loss": 0.4748, "mean_token_accuracy": 0.8437049388885498, "num_tokens": 83043763.0, "step": 2172 }, { "epoch": 0.27642793537717847, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.657227516174316, "learning_rate": 9.20729122509538e-07, "loss": 0.4176, "mean_token_accuracy": 0.8633748888969421, "num_tokens": 83081389.0, "step": 2173 }, { "epoch": 0.276555145655769, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.659655570983887, "learning_rate": 9.211530309453158e-07, "loss": 0.4572, "mean_token_accuracy": 0.8496639132499695, "num_tokens": 83120547.0, "step": 2174 }, { "epoch": 0.27668235593435947, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.639209747314453, "learning_rate": 9.215769393810936e-07, "loss": 0.4627, "mean_token_accuracy": 0.8512465953826904, "num_tokens": 83157229.0, "step": 2175 }, { "epoch": 0.27680956621295, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.685807228088379, "learning_rate": 9.220008478168715e-07, "loss": 0.5433, "mean_token_accuracy": 0.8260853290557861, "num_tokens": 83198217.0, "step": 2176 }, { "epoch": 0.2769367764915405, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.660948753356934, "learning_rate": 9.224247562526494e-07, "loss": 0.4637, "mean_token_accuracy": 0.8480408191680908, "num_tokens": 83238750.0, "step": 2177 }, { "epoch": 0.277063986770131, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.662711143493652, "learning_rate": 9.228486646884273e-07, "loss": 0.4521, "mean_token_accuracy": 0.8513592481613159, "num_tokens": 83280138.0, "step": 2178 }, { "epoch": 0.2771911970487215, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 8.728272438049316, "learning_rate": 9.232725731242052e-07, "loss": 0.5047, "mean_token_accuracy": 0.835885226726532, "num_tokens": 83320633.0, "step": 2179 }, { "epoch": 0.27731840732731206, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.653761863708496, "learning_rate": 9.23696481559983e-07, "loss": 0.4877, "mean_token_accuracy": 0.8398518562316895, "num_tokens": 83358787.0, "step": 2180 }, { "epoch": 0.2774456176059026, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.69182014465332, "learning_rate": 9.24120389995761e-07, "loss": 0.4307, "mean_token_accuracy": 0.8595993518829346, "num_tokens": 83400609.0, "step": 2181 }, { "epoch": 0.27757282788449306, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.671122550964355, "learning_rate": 9.245442984315387e-07, "loss": 0.4475, "mean_token_accuracy": 0.853818416595459, "num_tokens": 83436058.0, "step": 2182 }, { "epoch": 0.2777000381630836, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.718121528625488, "learning_rate": 9.249682068673165e-07, "loss": 0.4814, "mean_token_accuracy": 0.8419735431671143, "num_tokens": 83474542.0, "step": 2183 }, { "epoch": 0.2778272484416741, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.661101341247559, "learning_rate": 9.253921153030945e-07, "loss": 0.5172, "mean_token_accuracy": 0.8329868912696838, "num_tokens": 83515551.0, "step": 2184 }, { "epoch": 0.2779544587202646, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.68521785736084, "learning_rate": 9.258160237388723e-07, "loss": 0.4328, "mean_token_accuracy": 0.8606473207473755, "num_tokens": 83551826.0, "step": 2185 }, { "epoch": 0.2780816689988551, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.724259376525879, "learning_rate": 9.262399321746503e-07, "loss": 0.4502, "mean_token_accuracy": 0.8532559275627136, "num_tokens": 83586598.0, "step": 2186 }, { "epoch": 0.27820887927744564, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.703119277954102, "learning_rate": 9.266638406104281e-07, "loss": 0.4998, "mean_token_accuracy": 0.8385098576545715, "num_tokens": 83625996.0, "step": 2187 }, { "epoch": 0.2783360895560361, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.720165252685547, "learning_rate": 9.27087749046206e-07, "loss": 0.4687, "mean_token_accuracy": 0.8503887057304382, "num_tokens": 83657322.0, "step": 2188 }, { "epoch": 0.27846329983462664, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.710288047790527, "learning_rate": 9.275116574819839e-07, "loss": 0.464, "mean_token_accuracy": 0.8487547636032104, "num_tokens": 83693578.0, "step": 2189 }, { "epoch": 0.27859051011321717, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.74828815460205, "learning_rate": 9.279355659177617e-07, "loss": 0.4695, "mean_token_accuracy": 0.8496543169021606, "num_tokens": 83727672.0, "step": 2190 }, { "epoch": 0.27871772039180764, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.759950637817383, "learning_rate": 9.283594743535395e-07, "loss": 0.5131, "mean_token_accuracy": 0.8366180658340454, "num_tokens": 83761405.0, "step": 2191 }, { "epoch": 0.2788449306703982, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.692915916442871, "learning_rate": 9.287833827893175e-07, "loss": 0.4527, "mean_token_accuracy": 0.8489366769790649, "num_tokens": 83801564.0, "step": 2192 }, { "epoch": 0.2789721409489887, "ewc_loss": 0.010986328125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 8.736237525939941, "learning_rate": 9.292072912250953e-07, "loss": 0.4813, "mean_token_accuracy": 0.8434664607048035, "num_tokens": 83834260.0, "step": 2193 }, { "epoch": 0.2790993512275792, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.676410675048828, "learning_rate": 9.296311996608733e-07, "loss": 0.4099, "mean_token_accuracy": 0.8651157021522522, "num_tokens": 83873518.0, "step": 2194 }, { "epoch": 0.2792265615061697, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 8.721797943115234, "learning_rate": 9.300551080966511e-07, "loss": 0.524, "mean_token_accuracy": 0.8319355249404907, "num_tokens": 83913902.0, "step": 2195 }, { "epoch": 0.27935377178476023, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.724151611328125, "learning_rate": 9.30479016532429e-07, "loss": 0.45, "mean_token_accuracy": 0.8509209156036377, "num_tokens": 83947952.0, "step": 2196 }, { "epoch": 0.2794809820633507, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.726886749267578, "learning_rate": 9.309029249682068e-07, "loss": 0.4588, "mean_token_accuracy": 0.8520156741142273, "num_tokens": 83986057.0, "step": 2197 }, { "epoch": 0.27960819234194123, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.707419395446777, "learning_rate": 9.313268334039847e-07, "loss": 0.4692, "mean_token_accuracy": 0.8479020595550537, "num_tokens": 84023881.0, "step": 2198 }, { "epoch": 0.27973540262053176, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.70194149017334, "learning_rate": 9.317507418397625e-07, "loss": 0.4036, "mean_token_accuracy": 0.8687689900398254, "num_tokens": 84061480.0, "step": 2199 }, { "epoch": 0.27986261289912223, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.704524993896484, "learning_rate": 9.321746502755404e-07, "loss": 0.5051, "mean_token_accuracy": 0.838651716709137, "num_tokens": 84101782.0, "step": 2200 }, { "epoch": 0.27998982317771276, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.775175094604492, "learning_rate": 9.325985587113183e-07, "loss": 0.464, "mean_token_accuracy": 0.8491206169128418, "num_tokens": 84136125.0, "step": 2201 }, { "epoch": 0.2801170334563033, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.724752426147461, "learning_rate": 9.330224671470962e-07, "loss": 0.4766, "mean_token_accuracy": 0.850695013999939, "num_tokens": 84169748.0, "step": 2202 }, { "epoch": 0.28024424373489376, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.791553497314453, "learning_rate": 9.334463755828741e-07, "loss": 0.4914, "mean_token_accuracy": 0.836213231086731, "num_tokens": 84201292.0, "step": 2203 }, { "epoch": 0.2803714540134843, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.705946922302246, "learning_rate": 9.338702840186519e-07, "loss": 0.4912, "mean_token_accuracy": 0.8389157056808472, "num_tokens": 84240652.0, "step": 2204 }, { "epoch": 0.2804986642920748, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.737608909606934, "learning_rate": 9.342941924544298e-07, "loss": 0.4826, "mean_token_accuracy": 0.8437788486480713, "num_tokens": 84280252.0, "step": 2205 }, { "epoch": 0.2806258745706653, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.74658489227295, "learning_rate": 9.347181008902076e-07, "loss": 0.4214, "mean_token_accuracy": 0.8653963208198547, "num_tokens": 84316505.0, "step": 2206 }, { "epoch": 0.2807530848492558, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.737432479858398, "learning_rate": 9.351420093259855e-07, "loss": 0.4245, "mean_token_accuracy": 0.8626396656036377, "num_tokens": 84349990.0, "step": 2207 }, { "epoch": 0.28088029512784635, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.800287246704102, "learning_rate": 9.355659177617634e-07, "loss": 0.4144, "mean_token_accuracy": 0.8624278903007507, "num_tokens": 84381066.0, "step": 2208 }, { "epoch": 0.2810075054064368, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.669352531433105, "learning_rate": 9.359898261975413e-07, "loss": 0.4064, "mean_token_accuracy": 0.8661516904830933, "num_tokens": 84422540.0, "step": 2209 }, { "epoch": 0.28113471568502735, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 8.77144718170166, "learning_rate": 9.364137346333192e-07, "loss": 0.5339, "mean_token_accuracy": 0.8291760683059692, "num_tokens": 84462158.0, "step": 2210 }, { "epoch": 0.2812619259636179, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.779465675354004, "learning_rate": 9.368376430690971e-07, "loss": 0.4237, "mean_token_accuracy": 0.8606151938438416, "num_tokens": 84495887.0, "step": 2211 }, { "epoch": 0.28138913624220835, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.779293060302734, "learning_rate": 9.372615515048749e-07, "loss": 0.4957, "mean_token_accuracy": 0.8416321277618408, "num_tokens": 84530176.0, "step": 2212 }, { "epoch": 0.2815163465207989, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.72894287109375, "learning_rate": 9.376854599406528e-07, "loss": 0.4974, "mean_token_accuracy": 0.8412171602249146, "num_tokens": 84574450.0, "step": 2213 }, { "epoch": 0.2816435567993894, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.75229549407959, "learning_rate": 9.381093683764306e-07, "loss": 0.4651, "mean_token_accuracy": 0.8497914671897888, "num_tokens": 84610661.0, "step": 2214 }, { "epoch": 0.2817707670779799, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.748703002929688, "learning_rate": 9.385332768122085e-07, "loss": 0.4467, "mean_token_accuracy": 0.8534843325614929, "num_tokens": 84651259.0, "step": 2215 }, { "epoch": 0.2818979773565704, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.751959800720215, "learning_rate": 9.389571852479864e-07, "loss": 0.4508, "mean_token_accuracy": 0.8544647097587585, "num_tokens": 84692389.0, "step": 2216 }, { "epoch": 0.28202518763516093, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.799825668334961, "learning_rate": 9.393810936837643e-07, "loss": 0.4547, "mean_token_accuracy": 0.852149486541748, "num_tokens": 84731905.0, "step": 2217 }, { "epoch": 0.2821523979137514, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.737435340881348, "learning_rate": 9.398050021195422e-07, "loss": 0.4828, "mean_token_accuracy": 0.8475097417831421, "num_tokens": 84773727.0, "step": 2218 }, { "epoch": 0.28227960819234194, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.743484497070312, "learning_rate": 9.402289105553201e-07, "loss": 0.4288, "mean_token_accuracy": 0.8587803244590759, "num_tokens": 84816771.0, "step": 2219 }, { "epoch": 0.28240681847093246, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.756155967712402, "learning_rate": 9.406528189910978e-07, "loss": 0.4528, "mean_token_accuracy": 0.8534829020500183, "num_tokens": 84853172.0, "step": 2220 }, { "epoch": 0.28253402874952294, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.781340599060059, "learning_rate": 9.410767274268757e-07, "loss": 0.4334, "mean_token_accuracy": 0.8567098379135132, "num_tokens": 84891508.0, "step": 2221 }, { "epoch": 0.28266123902811346, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.861764907836914, "learning_rate": 9.415006358626536e-07, "loss": 0.5312, "mean_token_accuracy": 0.8277996778488159, "num_tokens": 84931204.0, "step": 2222 }, { "epoch": 0.282788449306704, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.8380126953125, "learning_rate": 9.419245442984314e-07, "loss": 0.5126, "mean_token_accuracy": 0.8353840112686157, "num_tokens": 84966782.0, "step": 2223 }, { "epoch": 0.28291565958529447, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.826377868652344, "learning_rate": 9.423484527342094e-07, "loss": 0.4582, "mean_token_accuracy": 0.848127007484436, "num_tokens": 85001512.0, "step": 2224 }, { "epoch": 0.283042869863885, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.812525749206543, "learning_rate": 9.427723611699872e-07, "loss": 0.4665, "mean_token_accuracy": 0.8504587411880493, "num_tokens": 85036082.0, "step": 2225 }, { "epoch": 0.2831700801424755, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.71517562866211, "learning_rate": 9.431962696057652e-07, "loss": 0.4735, "mean_token_accuracy": 0.8471406698226929, "num_tokens": 85079433.0, "step": 2226 }, { "epoch": 0.283297290421066, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.816388130187988, "learning_rate": 9.43620178041543e-07, "loss": 0.4246, "mean_token_accuracy": 0.8633190393447876, "num_tokens": 85116940.0, "step": 2227 }, { "epoch": 0.2834245006996565, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.75784969329834, "learning_rate": 9.440440864773208e-07, "loss": 0.4779, "mean_token_accuracy": 0.8473907709121704, "num_tokens": 85153328.0, "step": 2228 }, { "epoch": 0.28355171097824705, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.74141788482666, "learning_rate": 9.444679949130987e-07, "loss": 0.4794, "mean_token_accuracy": 0.8431888222694397, "num_tokens": 85194059.0, "step": 2229 }, { "epoch": 0.2836789212568376, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.84598445892334, "learning_rate": 9.448919033488766e-07, "loss": 0.4992, "mean_token_accuracy": 0.8420270681381226, "num_tokens": 85230820.0, "step": 2230 }, { "epoch": 0.28380613153542805, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.746238708496094, "learning_rate": 9.453158117846544e-07, "loss": 0.481, "mean_token_accuracy": 0.8450702428817749, "num_tokens": 85268864.0, "step": 2231 }, { "epoch": 0.2839333418140186, "ewc_loss": 0.01123046875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 8.772557258605957, "learning_rate": 9.457397202204324e-07, "loss": 0.5072, "mean_token_accuracy": 0.8376773595809937, "num_tokens": 85311377.0, "step": 2232 }, { "epoch": 0.2840605520926091, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.79263687133789, "learning_rate": 9.461636286562102e-07, "loss": 0.422, "mean_token_accuracy": 0.8644516468048096, "num_tokens": 85347082.0, "step": 2233 }, { "epoch": 0.2841877623711996, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.827249526977539, "learning_rate": 9.465875370919882e-07, "loss": 0.51, "mean_token_accuracy": 0.8394647240638733, "num_tokens": 85384696.0, "step": 2234 }, { "epoch": 0.2843149726497901, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.824379920959473, "learning_rate": 9.470114455277659e-07, "loss": 0.4961, "mean_token_accuracy": 0.8392763137817383, "num_tokens": 85420192.0, "step": 2235 }, { "epoch": 0.28444218292838064, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.782596588134766, "learning_rate": 9.474353539635438e-07, "loss": 0.4464, "mean_token_accuracy": 0.8536602854728699, "num_tokens": 85462281.0, "step": 2236 }, { "epoch": 0.2845693932069711, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.838531494140625, "learning_rate": 9.478592623993217e-07, "loss": 0.4521, "mean_token_accuracy": 0.8516424298286438, "num_tokens": 85495191.0, "step": 2237 }, { "epoch": 0.28469660348556164, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.856453895568848, "learning_rate": 9.482831708350996e-07, "loss": 0.45, "mean_token_accuracy": 0.8551108241081238, "num_tokens": 85533803.0, "step": 2238 }, { "epoch": 0.28482381376415217, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.802818298339844, "learning_rate": 9.487070792708775e-07, "loss": 0.4963, "mean_token_accuracy": 0.8421893119812012, "num_tokens": 85573424.0, "step": 2239 }, { "epoch": 0.28495102404274264, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.821958541870117, "learning_rate": 9.491309877066554e-07, "loss": 0.4798, "mean_token_accuracy": 0.8447871208190918, "num_tokens": 85616423.0, "step": 2240 }, { "epoch": 0.28507823432133317, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.849085807800293, "learning_rate": 9.495548961424332e-07, "loss": 0.5053, "mean_token_accuracy": 0.8371464014053345, "num_tokens": 85658410.0, "step": 2241 }, { "epoch": 0.2852054445999237, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.772639274597168, "learning_rate": 9.499788045782111e-07, "loss": 0.4635, "mean_token_accuracy": 0.8494915962219238, "num_tokens": 85695767.0, "step": 2242 }, { "epoch": 0.28533265487851417, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.831616401672363, "learning_rate": 9.504027130139889e-07, "loss": 0.4791, "mean_token_accuracy": 0.8444614410400391, "num_tokens": 85737771.0, "step": 2243 }, { "epoch": 0.2854598651571047, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.76651668548584, "learning_rate": 9.508266214497667e-07, "loss": 0.4265, "mean_token_accuracy": 0.8639233708381653, "num_tokens": 85774789.0, "step": 2244 }, { "epoch": 0.2855870754356952, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 8.80684757232666, "learning_rate": 9.512505298855447e-07, "loss": 0.4653, "mean_token_accuracy": 0.8513696193695068, "num_tokens": 85817833.0, "step": 2245 }, { "epoch": 0.2857142857142857, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.836236000061035, "learning_rate": 9.516744383213225e-07, "loss": 0.4092, "mean_token_accuracy": 0.8651592135429382, "num_tokens": 85854488.0, "step": 2246 }, { "epoch": 0.2858414959928762, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.808746337890625, "learning_rate": 9.520983467571005e-07, "loss": 0.4875, "mean_token_accuracy": 0.8437192440032959, "num_tokens": 85893746.0, "step": 2247 }, { "epoch": 0.28596870627146675, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 8.850186347961426, "learning_rate": 9.525222551928783e-07, "loss": 0.5268, "mean_token_accuracy": 0.8366252183914185, "num_tokens": 85932798.0, "step": 2248 }, { "epoch": 0.2860959165500572, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.86702823638916, "learning_rate": 9.529461636286562e-07, "loss": 0.4405, "mean_token_accuracy": 0.8566666841506958, "num_tokens": 85964446.0, "step": 2249 }, { "epoch": 0.28622312682864776, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.81514835357666, "learning_rate": 9.533700720644341e-07, "loss": 0.4184, "mean_token_accuracy": 0.8650311231613159, "num_tokens": 86000587.0, "step": 2250 }, { "epoch": 0.2863503371072383, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.812318801879883, "learning_rate": 9.537939805002118e-07, "loss": 0.4708, "mean_token_accuracy": 0.8478327393531799, "num_tokens": 86042891.0, "step": 2251 }, { "epoch": 0.28647754738582876, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.836395263671875, "learning_rate": 9.542178889359898e-07, "loss": 0.4485, "mean_token_accuracy": 0.8548266887664795, "num_tokens": 86079973.0, "step": 2252 }, { "epoch": 0.2866047576644193, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.82779312133789, "learning_rate": 9.546417973717677e-07, "loss": 0.4714, "mean_token_accuracy": 0.8493772745132446, "num_tokens": 86119767.0, "step": 2253 }, { "epoch": 0.2867319679430098, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.822017669677734, "learning_rate": 9.550657058075455e-07, "loss": 0.4914, "mean_token_accuracy": 0.843114972114563, "num_tokens": 86163176.0, "step": 2254 }, { "epoch": 0.2868591782216003, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.825676918029785, "learning_rate": 9.554896142433234e-07, "loss": 0.486, "mean_token_accuracy": 0.8425494432449341, "num_tokens": 86201523.0, "step": 2255 }, { "epoch": 0.2869863885001908, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.880462646484375, "learning_rate": 9.559135226791012e-07, "loss": 0.5049, "mean_token_accuracy": 0.8336424827575684, "num_tokens": 86240277.0, "step": 2256 }, { "epoch": 0.28711359877878134, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.832526206970215, "learning_rate": 9.563374311148793e-07, "loss": 0.4952, "mean_token_accuracy": 0.8404738903045654, "num_tokens": 86289714.0, "step": 2257 }, { "epoch": 0.2872408090573718, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.856375694274902, "learning_rate": 9.56761339550657e-07, "loss": 0.4467, "mean_token_accuracy": 0.8567649126052856, "num_tokens": 86320786.0, "step": 2258 }, { "epoch": 0.28736801933596234, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.811328887939453, "learning_rate": 9.57185247986435e-07, "loss": 0.4358, "mean_token_accuracy": 0.8585599660873413, "num_tokens": 86358002.0, "step": 2259 }, { "epoch": 0.28749522961455287, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.89944076538086, "learning_rate": 9.576091564222128e-07, "loss": 0.4042, "mean_token_accuracy": 0.8659411668777466, "num_tokens": 86395334.0, "step": 2260 }, { "epoch": 0.28762243989314334, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.808518409729004, "learning_rate": 9.580330648579906e-07, "loss": 0.482, "mean_token_accuracy": 0.8418501615524292, "num_tokens": 86432616.0, "step": 2261 }, { "epoch": 0.2877496501717339, "ewc_loss": 0.011474609375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 8.822978019714355, "learning_rate": 9.584569732937685e-07, "loss": 0.4765, "mean_token_accuracy": 0.8446950316429138, "num_tokens": 86472293.0, "step": 2262 }, { "epoch": 0.2878768604503244, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.843727111816406, "learning_rate": 9.588808817295463e-07, "loss": 0.4662, "mean_token_accuracy": 0.8456525802612305, "num_tokens": 86506304.0, "step": 2263 }, { "epoch": 0.2880040707289149, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.852787971496582, "learning_rate": 9.593047901653242e-07, "loss": 0.4648, "mean_token_accuracy": 0.8476914167404175, "num_tokens": 86540520.0, "step": 2264 }, { "epoch": 0.2881312810075054, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 8.818788528442383, "learning_rate": 9.597286986011022e-07, "loss": 0.425, "mean_token_accuracy": 0.8614380359649658, "num_tokens": 86576956.0, "step": 2265 }, { "epoch": 0.28825849128609593, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.861967086791992, "learning_rate": 9.601526070368799e-07, "loss": 0.5097, "mean_token_accuracy": 0.8366757035255432, "num_tokens": 86613799.0, "step": 2266 }, { "epoch": 0.2883857015646864, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.86039924621582, "learning_rate": 9.60576515472658e-07, "loss": 0.4499, "mean_token_accuracy": 0.8539527654647827, "num_tokens": 86650338.0, "step": 2267 }, { "epoch": 0.28851291184327693, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.868653297424316, "learning_rate": 9.610004239084358e-07, "loss": 0.4616, "mean_token_accuracy": 0.8490816354751587, "num_tokens": 86691605.0, "step": 2268 }, { "epoch": 0.28864012212186746, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.855195045471191, "learning_rate": 9.614243323442136e-07, "loss": 0.53, "mean_token_accuracy": 0.8281846046447754, "num_tokens": 86728767.0, "step": 2269 }, { "epoch": 0.28876733240045793, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.938427925109863, "learning_rate": 9.618482407799915e-07, "loss": 0.4419, "mean_token_accuracy": 0.8542152643203735, "num_tokens": 86765240.0, "step": 2270 }, { "epoch": 0.28889454267904846, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.868285179138184, "learning_rate": 9.622721492157693e-07, "loss": 0.4101, "mean_token_accuracy": 0.8669112324714661, "num_tokens": 86804042.0, "step": 2271 }, { "epoch": 0.289021752957639, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.920942306518555, "learning_rate": 9.626960576515472e-07, "loss": 0.4604, "mean_token_accuracy": 0.8507983684539795, "num_tokens": 86835220.0, "step": 2272 }, { "epoch": 0.28914896323622946, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.881552696228027, "learning_rate": 9.63119966087325e-07, "loss": 0.4464, "mean_token_accuracy": 0.852104127407074, "num_tokens": 86873414.0, "step": 2273 }, { "epoch": 0.28927617351482, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.8619384765625, "learning_rate": 9.635438745231029e-07, "loss": 0.4372, "mean_token_accuracy": 0.8587080240249634, "num_tokens": 86914272.0, "step": 2274 }, { "epoch": 0.2894033837934105, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.943841934204102, "learning_rate": 9.63967782958881e-07, "loss": 0.4776, "mean_token_accuracy": 0.8426718711853027, "num_tokens": 86952425.0, "step": 2275 }, { "epoch": 0.289530594072001, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.89276123046875, "learning_rate": 9.643916913946588e-07, "loss": 0.437, "mean_token_accuracy": 0.8581894636154175, "num_tokens": 86987842.0, "step": 2276 }, { "epoch": 0.2896578043505915, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.873808860778809, "learning_rate": 9.648155998304366e-07, "loss": 0.4104, "mean_token_accuracy": 0.8663539886474609, "num_tokens": 87026750.0, "step": 2277 }, { "epoch": 0.28978501462918205, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.965993881225586, "learning_rate": 9.652395082662145e-07, "loss": 0.5141, "mean_token_accuracy": 0.8326611518859863, "num_tokens": 87069685.0, "step": 2278 }, { "epoch": 0.2899122249077726, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.913832664489746, "learning_rate": 9.656634167019923e-07, "loss": 0.4694, "mean_token_accuracy": 0.8439009189605713, "num_tokens": 87107054.0, "step": 2279 }, { "epoch": 0.29003943518636305, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.984601020812988, "learning_rate": 9.660873251377701e-07, "loss": 0.4759, "mean_token_accuracy": 0.8461799621582031, "num_tokens": 87146254.0, "step": 2280 }, { "epoch": 0.2901666454649536, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.903898239135742, "learning_rate": 9.66511233573548e-07, "loss": 0.5011, "mean_token_accuracy": 0.842242956161499, "num_tokens": 87182404.0, "step": 2281 }, { "epoch": 0.2902938557435441, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.895421981811523, "learning_rate": 9.669351420093258e-07, "loss": 0.4326, "mean_token_accuracy": 0.8592469692230225, "num_tokens": 87216551.0, "step": 2282 }, { "epoch": 0.2904210660221346, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.980180740356445, "learning_rate": 9.67359050445104e-07, "loss": 0.456, "mean_token_accuracy": 0.8504166007041931, "num_tokens": 87249402.0, "step": 2283 }, { "epoch": 0.2905482763007251, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.895099639892578, "learning_rate": 9.677829588808817e-07, "loss": 0.4434, "mean_token_accuracy": 0.8522941470146179, "num_tokens": 87281846.0, "step": 2284 }, { "epoch": 0.29067548657931563, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.924338340759277, "learning_rate": 9.682068673166596e-07, "loss": 0.5134, "mean_token_accuracy": 0.8386916518211365, "num_tokens": 87315765.0, "step": 2285 }, { "epoch": 0.2908026968579061, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.983963966369629, "learning_rate": 9.686307757524374e-07, "loss": 0.4567, "mean_token_accuracy": 0.850480318069458, "num_tokens": 87353668.0, "step": 2286 }, { "epoch": 0.29092990713649663, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.885486602783203, "learning_rate": 9.690546841882153e-07, "loss": 0.3883, "mean_token_accuracy": 0.8731745481491089, "num_tokens": 87390747.0, "step": 2287 }, { "epoch": 0.29105711741508716, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.926004409790039, "learning_rate": 9.694785926239931e-07, "loss": 0.469, "mean_token_accuracy": 0.850540280342102, "num_tokens": 87430686.0, "step": 2288 }, { "epoch": 0.29118432769367764, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.933066368103027, "learning_rate": 9.69902501059771e-07, "loss": 0.5082, "mean_token_accuracy": 0.8354794979095459, "num_tokens": 87465290.0, "step": 2289 }, { "epoch": 0.29131153797226816, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.920868873596191, "learning_rate": 9.703264094955488e-07, "loss": 0.4781, "mean_token_accuracy": 0.8447480797767639, "num_tokens": 87504553.0, "step": 2290 }, { "epoch": 0.2914387482508587, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.935543060302734, "learning_rate": 9.707503179313269e-07, "loss": 0.4476, "mean_token_accuracy": 0.8563770055770874, "num_tokens": 87544896.0, "step": 2291 }, { "epoch": 0.29156595852944917, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.96910572052002, "learning_rate": 9.711742263671047e-07, "loss": 0.4479, "mean_token_accuracy": 0.8543175458908081, "num_tokens": 87582075.0, "step": 2292 }, { "epoch": 0.2916931688080397, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.964204788208008, "learning_rate": 9.715981348028826e-07, "loss": 0.441, "mean_token_accuracy": 0.8560612797737122, "num_tokens": 87623117.0, "step": 2293 }, { "epoch": 0.2918203790866302, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.994035720825195, "learning_rate": 9.720220432386604e-07, "loss": 0.4555, "mean_token_accuracy": 0.8540773391723633, "num_tokens": 87656263.0, "step": 2294 }, { "epoch": 0.2919475893652207, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.929078102111816, "learning_rate": 9.724459516744383e-07, "loss": 0.4294, "mean_token_accuracy": 0.8604130744934082, "num_tokens": 87695852.0, "step": 2295 }, { "epoch": 0.2920747996438112, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.954333305358887, "learning_rate": 9.728698601102161e-07, "loss": 0.4274, "mean_token_accuracy": 0.8608129024505615, "num_tokens": 87734331.0, "step": 2296 }, { "epoch": 0.29220200992240175, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.964210510253906, "learning_rate": 9.73293768545994e-07, "loss": 0.4607, "mean_token_accuracy": 0.8471764326095581, "num_tokens": 87768447.0, "step": 2297 }, { "epoch": 0.2923292202009922, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.964059829711914, "learning_rate": 9.737176769817718e-07, "loss": 0.4884, "mean_token_accuracy": 0.846865177154541, "num_tokens": 87813213.0, "step": 2298 }, { "epoch": 0.29245643047958275, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.976725578308105, "learning_rate": 9.741415854175499e-07, "loss": 0.5239, "mean_token_accuracy": 0.8359971046447754, "num_tokens": 87851336.0, "step": 2299 }, { "epoch": 0.2925836407581733, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.961668014526367, "learning_rate": 9.745654938533277e-07, "loss": 0.4774, "mean_token_accuracy": 0.8457112312316895, "num_tokens": 87888474.0, "step": 2300 }, { "epoch": 0.29271085103676375, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 8.98137378692627, "learning_rate": 9.749894022891056e-07, "loss": 0.5026, "mean_token_accuracy": 0.8394719362258911, "num_tokens": 87928232.0, "step": 2301 }, { "epoch": 0.2928380613153543, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.93208122253418, "learning_rate": 9.754133107248834e-07, "loss": 0.4203, "mean_token_accuracy": 0.8614223599433899, "num_tokens": 87966982.0, "step": 2302 }, { "epoch": 0.2929652715939448, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.957447052001953, "learning_rate": 9.758372191606612e-07, "loss": 0.5067, "mean_token_accuracy": 0.8354828357696533, "num_tokens": 88003550.0, "step": 2303 }, { "epoch": 0.2930924818725353, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 8.971516609191895, "learning_rate": 9.76261127596439e-07, "loss": 0.4175, "mean_token_accuracy": 0.8646941184997559, "num_tokens": 88038038.0, "step": 2304 }, { "epoch": 0.2932196921511258, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.977457046508789, "learning_rate": 9.76685036032217e-07, "loss": 0.489, "mean_token_accuracy": 0.8390519618988037, "num_tokens": 88074627.0, "step": 2305 }, { "epoch": 0.29334690242971634, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.955229759216309, "learning_rate": 9.771089444679948e-07, "loss": 0.4326, "mean_token_accuracy": 0.8587679862976074, "num_tokens": 88116772.0, "step": 2306 }, { "epoch": 0.2934741127083068, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 9.004666328430176, "learning_rate": 9.775328529037728e-07, "loss": 0.4774, "mean_token_accuracy": 0.8457093238830566, "num_tokens": 88152888.0, "step": 2307 }, { "epoch": 0.29360132298689734, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 9.029365539550781, "learning_rate": 9.779567613395507e-07, "loss": 0.5067, "mean_token_accuracy": 0.8393509984016418, "num_tokens": 88184544.0, "step": 2308 }, { "epoch": 0.29372853326548787, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.974761009216309, "learning_rate": 9.783806697753285e-07, "loss": 0.4501, "mean_token_accuracy": 0.8549422025680542, "num_tokens": 88221603.0, "step": 2309 }, { "epoch": 0.29385574354407834, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.930150985717773, "learning_rate": 9.788045782111064e-07, "loss": 0.4248, "mean_token_accuracy": 0.8611850738525391, "num_tokens": 88265109.0, "step": 2310 }, { "epoch": 0.29398295382266887, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.990975379943848, "learning_rate": 9.792284866468842e-07, "loss": 0.4158, "mean_token_accuracy": 0.8614234924316406, "num_tokens": 88302756.0, "step": 2311 }, { "epoch": 0.2941101641012594, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.957355499267578, "learning_rate": 9.79652395082662e-07, "loss": 0.415, "mean_token_accuracy": 0.8587515950202942, "num_tokens": 88335605.0, "step": 2312 }, { "epoch": 0.29423737437984987, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 8.99197006225586, "learning_rate": 9.8007630351844e-07, "loss": 0.4851, "mean_token_accuracy": 0.8443317413330078, "num_tokens": 88369447.0, "step": 2313 }, { "epoch": 0.2943645846584404, "ewc_loss": 0.01171875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 8.980487823486328, "learning_rate": 9.805002119542178e-07, "loss": 0.515, "mean_token_accuracy": 0.833975613117218, "num_tokens": 88404442.0, "step": 2314 }, { "epoch": 0.2944917949370309, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.029505729675293, "learning_rate": 9.809241203899958e-07, "loss": 0.4607, "mean_token_accuracy": 0.8524131774902344, "num_tokens": 88445870.0, "step": 2315 }, { "epoch": 0.2946190052156214, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 8.96621322631836, "learning_rate": 9.813480288257737e-07, "loss": 0.4748, "mean_token_accuracy": 0.8415807485580444, "num_tokens": 88480994.0, "step": 2316 }, { "epoch": 0.2947462154942119, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.034823417663574, "learning_rate": 9.817719372615515e-07, "loss": 0.4309, "mean_token_accuracy": 0.8598788380622864, "num_tokens": 88519666.0, "step": 2317 }, { "epoch": 0.29487342577280246, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 8.976764678955078, "learning_rate": 9.821958456973294e-07, "loss": 0.4048, "mean_token_accuracy": 0.8654747009277344, "num_tokens": 88553855.0, "step": 2318 }, { "epoch": 0.29500063605139293, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.034536361694336, "learning_rate": 9.826197541331072e-07, "loss": 0.4205, "mean_token_accuracy": 0.8636878728866577, "num_tokens": 88589053.0, "step": 2319 }, { "epoch": 0.29512784632998346, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.024353981018066, "learning_rate": 9.83043662568885e-07, "loss": 0.4788, "mean_token_accuracy": 0.8456953167915344, "num_tokens": 88627067.0, "step": 2320 }, { "epoch": 0.295255056608574, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.949723243713379, "learning_rate": 9.83467571004663e-07, "loss": 0.4231, "mean_token_accuracy": 0.8611371517181396, "num_tokens": 88668030.0, "step": 2321 }, { "epoch": 0.29538226688716446, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.006024360656738, "learning_rate": 9.838914794404407e-07, "loss": 0.4215, "mean_token_accuracy": 0.8635784983634949, "num_tokens": 88708889.0, "step": 2322 }, { "epoch": 0.295509477165755, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.073494911193848, "learning_rate": 9.843153878762188e-07, "loss": 0.4885, "mean_token_accuracy": 0.844059407711029, "num_tokens": 88745689.0, "step": 2323 }, { "epoch": 0.2956366874443455, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 8.961196899414062, "learning_rate": 9.847392963119966e-07, "loss": 0.4405, "mean_token_accuracy": 0.857944905757904, "num_tokens": 88784667.0, "step": 2324 }, { "epoch": 0.295763897722936, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.002740859985352, "learning_rate": 9.851632047477745e-07, "loss": 0.4533, "mean_token_accuracy": 0.8520318865776062, "num_tokens": 88827334.0, "step": 2325 }, { "epoch": 0.2958911080015265, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.003419876098633, "learning_rate": 9.855871131835523e-07, "loss": 0.3868, "mean_token_accuracy": 0.8722906112670898, "num_tokens": 88863744.0, "step": 2326 }, { "epoch": 0.29601831828011704, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.005485534667969, "learning_rate": 9.860110216193302e-07, "loss": 0.4189, "mean_token_accuracy": 0.8645657300949097, "num_tokens": 88903693.0, "step": 2327 }, { "epoch": 0.2961455285587075, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.036688804626465, "learning_rate": 9.86434930055108e-07, "loss": 0.4146, "mean_token_accuracy": 0.8643366098403931, "num_tokens": 88936668.0, "step": 2328 }, { "epoch": 0.29627273883729804, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.053040504455566, "learning_rate": 9.868588384908859e-07, "loss": 0.4238, "mean_token_accuracy": 0.8582533001899719, "num_tokens": 88970920.0, "step": 2329 }, { "epoch": 0.2963999491158886, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.054183006286621, "learning_rate": 9.872827469266637e-07, "loss": 0.4269, "mean_token_accuracy": 0.858150064945221, "num_tokens": 89010856.0, "step": 2330 }, { "epoch": 0.2965271593944791, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.087282180786133, "learning_rate": 9.877066553624418e-07, "loss": 0.4354, "mean_token_accuracy": 0.8572174906730652, "num_tokens": 89047109.0, "step": 2331 }, { "epoch": 0.2966543696730696, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.032865524291992, "learning_rate": 9.881305637982196e-07, "loss": 0.4438, "mean_token_accuracy": 0.8564585447311401, "num_tokens": 89087859.0, "step": 2332 }, { "epoch": 0.2967815799516601, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.040064811706543, "learning_rate": 9.885544722339975e-07, "loss": 0.426, "mean_token_accuracy": 0.8588725328445435, "num_tokens": 89123601.0, "step": 2333 }, { "epoch": 0.29690879023025063, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.053083419799805, "learning_rate": 9.889783806697753e-07, "loss": 0.4087, "mean_token_accuracy": 0.8674050569534302, "num_tokens": 89160133.0, "step": 2334 }, { "epoch": 0.2970360005088411, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.051219940185547, "learning_rate": 9.894022891055532e-07, "loss": 0.4032, "mean_token_accuracy": 0.8690735101699829, "num_tokens": 89196731.0, "step": 2335 }, { "epoch": 0.29716321078743163, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.07040786743164, "learning_rate": 9.89826197541331e-07, "loss": 0.3919, "mean_token_accuracy": 0.8717544674873352, "num_tokens": 89230159.0, "step": 2336 }, { "epoch": 0.29729042106602216, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 9.111252784729004, "learning_rate": 9.902501059771089e-07, "loss": 0.4561, "mean_token_accuracy": 0.8503903150558472, "num_tokens": 89264648.0, "step": 2337 }, { "epoch": 0.29741763134461263, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.021027565002441, "learning_rate": 9.906740144128867e-07, "loss": 0.4552, "mean_token_accuracy": 0.8495588302612305, "num_tokens": 89302505.0, "step": 2338 }, { "epoch": 0.29754484162320316, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.122946739196777, "learning_rate": 9.910979228486648e-07, "loss": 0.4144, "mean_token_accuracy": 0.8626346588134766, "num_tokens": 89338079.0, "step": 2339 }, { "epoch": 0.2976720519017937, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.058297157287598, "learning_rate": 9.915218312844426e-07, "loss": 0.4794, "mean_token_accuracy": 0.8494819402694702, "num_tokens": 89376717.0, "step": 2340 }, { "epoch": 0.29779926218038416, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.096980094909668, "learning_rate": 9.919457397202205e-07, "loss": 0.4692, "mean_token_accuracy": 0.8474583625793457, "num_tokens": 89416164.0, "step": 2341 }, { "epoch": 0.2979264724589747, "ewc_loss": 0.011962890625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 9.06695556640625, "learning_rate": 9.923696481559983e-07, "loss": 0.4241, "mean_token_accuracy": 0.8621706962585449, "num_tokens": 89446491.0, "step": 2342 }, { "epoch": 0.2980536827375652, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.092462539672852, "learning_rate": 9.927935565917761e-07, "loss": 0.4716, "mean_token_accuracy": 0.84449702501297, "num_tokens": 89483637.0, "step": 2343 }, { "epoch": 0.2981808930161557, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.045750617980957, "learning_rate": 9.93217465027554e-07, "loss": 0.4641, "mean_token_accuracy": 0.8530628085136414, "num_tokens": 89526532.0, "step": 2344 }, { "epoch": 0.2983081032947462, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.119786262512207, "learning_rate": 9.936413734633318e-07, "loss": 0.5255, "mean_token_accuracy": 0.8361361026763916, "num_tokens": 89558353.0, "step": 2345 }, { "epoch": 0.29843531357333675, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.075057983398438, "learning_rate": 9.940652818991097e-07, "loss": 0.4176, "mean_token_accuracy": 0.8643549084663391, "num_tokens": 89598727.0, "step": 2346 }, { "epoch": 0.2985625238519272, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.038540840148926, "learning_rate": 9.944891903348877e-07, "loss": 0.485, "mean_token_accuracy": 0.8430800437927246, "num_tokens": 89640093.0, "step": 2347 }, { "epoch": 0.29868973413051775, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.068214416503906, "learning_rate": 9.949130987706656e-07, "loss": 0.4307, "mean_token_accuracy": 0.859465479850769, "num_tokens": 89681401.0, "step": 2348 }, { "epoch": 0.2988169444091083, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.062560081481934, "learning_rate": 9.953370072064432e-07, "loss": 0.4233, "mean_token_accuracy": 0.8603826761245728, "num_tokens": 89720366.0, "step": 2349 }, { "epoch": 0.29894415468769875, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.096529006958008, "learning_rate": 9.957609156422213e-07, "loss": 0.5007, "mean_token_accuracy": 0.8383687734603882, "num_tokens": 89762234.0, "step": 2350 }, { "epoch": 0.2990713649662893, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.14097785949707, "learning_rate": 9.961848240779991e-07, "loss": 0.4464, "mean_token_accuracy": 0.8554198741912842, "num_tokens": 89792208.0, "step": 2351 }, { "epoch": 0.2991985752448798, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.11043930053711, "learning_rate": 9.96608732513777e-07, "loss": 0.4248, "mean_token_accuracy": 0.8655350208282471, "num_tokens": 89833246.0, "step": 2352 }, { "epoch": 0.2993257855234703, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 9.028488159179688, "learning_rate": 9.970326409495548e-07, "loss": 0.483, "mean_token_accuracy": 0.8426761031150818, "num_tokens": 89873271.0, "step": 2353 }, { "epoch": 0.2994529958020608, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.11781120300293, "learning_rate": 9.974565493853327e-07, "loss": 0.4979, "mean_token_accuracy": 0.8381128311157227, "num_tokens": 89913069.0, "step": 2354 }, { "epoch": 0.29958020608065133, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.087923049926758, "learning_rate": 9.978804578211107e-07, "loss": 0.4787, "mean_token_accuracy": 0.8441555500030518, "num_tokens": 89948915.0, "step": 2355 }, { "epoch": 0.2997074163592418, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.138165473937988, "learning_rate": 9.983043662568886e-07, "loss": 0.4762, "mean_token_accuracy": 0.8473217487335205, "num_tokens": 89982123.0, "step": 2356 }, { "epoch": 0.29983462663783234, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.049562454223633, "learning_rate": 9.987282746926662e-07, "loss": 0.4278, "mean_token_accuracy": 0.8602322340011597, "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": 9.139320373535156, "learning_rate": 9.991521831284443e-07, "loss": 0.4481, "mean_token_accuracy": 0.8550277948379517, "num_tokens": 90059041.0, "step": 2358 }, { "epoch": 0.30008904719501334, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.121169090270996, "learning_rate": 9.995760915642221e-07, "loss": 0.3975, "mean_token_accuracy": 0.8699678182601929, "num_tokens": 90089463.0, "step": 2359 }, { "epoch": 0.30021625747360386, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.136312484741211, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.861442506313324, "num_tokens": 90125691.0, "step": 2360 }, { "epoch": 0.3003434677521944, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.095152854919434, "learning_rate": 1e-06, "loss": 0.4693, "mean_token_accuracy": 0.8512693047523499, "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": 9.089103698730469, "learning_rate": 1e-06, "loss": 0.4997, "mean_token_accuracy": 0.8392798900604248, "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": 9.079145431518555, "learning_rate": 1e-06, "loss": 0.4831, "mean_token_accuracy": 0.8452287316322327, "num_tokens": 90248924.0, "step": 2363 }, { "epoch": 0.3007250985879659, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.193081855773926, "learning_rate": 1e-06, "loss": 0.5651, "mean_token_accuracy": 0.8244144916534424, "num_tokens": 90281988.0, "step": 2364 }, { "epoch": 0.3008523088665564, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.115915298461914, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.846976637840271, "num_tokens": 90317050.0, "step": 2365 }, { "epoch": 0.3009795191451469, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.087671279907227, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.860714852809906, "num_tokens": 90358064.0, "step": 2366 }, { "epoch": 0.30110672942373745, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.11715316772461, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8408840894699097, "num_tokens": 90400692.0, "step": 2367 }, { "epoch": 0.3012339397023279, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.15720272064209, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8449965715408325, "num_tokens": 90432207.0, "step": 2368 }, { "epoch": 0.30136114998091845, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.16695499420166, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8501982688903809, "num_tokens": 90466081.0, "step": 2369 }, { "epoch": 0.301488360259509, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.140607833862305, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8594598770141602, "num_tokens": 90501963.0, "step": 2370 }, { "epoch": 0.30161557053809945, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.13305950164795, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8638616800308228, "num_tokens": 90540651.0, "step": 2371 }, { "epoch": 0.30174278081669, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.13477611541748, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8445504903793335, "num_tokens": 90585105.0, "step": 2372 }, { "epoch": 0.3018699910952805, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.105259895324707, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8491100072860718, "num_tokens": 90626773.0, "step": 2373 }, { "epoch": 0.301997201373871, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.117679595947266, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8662962317466736, "num_tokens": 90665974.0, "step": 2374 }, { "epoch": 0.3021244116524615, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.129081726074219, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8474715948104858, "num_tokens": 90706801.0, "step": 2375 }, { "epoch": 0.30225162193105204, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.159647941589355, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8486220240592957, "num_tokens": 90743635.0, "step": 2376 }, { "epoch": 0.3023788322096425, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.083803176879883, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8531407117843628, "num_tokens": 90782797.0, "step": 2377 }, { "epoch": 0.30250604248823304, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3126602172851562e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.140087127685547, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8445180654525757, "num_tokens": 90822661.0, "step": 2378 }, { "epoch": 0.30263325276682357, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.158721923828125, "learning_rate": 1e-06, "loss": 0.498, "mean_token_accuracy": 0.8343193531036377, "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": 9.14836597442627, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.85004723072052, "num_tokens": 90895778.0, "step": 2380 }, { "epoch": 0.30288767332400457, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 9.093489646911621, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8585755825042725, "num_tokens": 90934463.0, "step": 2381 }, { "epoch": 0.3030148836025951, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.11539363861084, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8569172620773315, "num_tokens": 90967888.0, "step": 2382 }, { "epoch": 0.3031420938811856, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.19192123413086, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8604257702827454, "num_tokens": 91005200.0, "step": 2383 }, { "epoch": 0.3032693041597761, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.136494636535645, "learning_rate": 1e-06, "loss": 0.5153, "mean_token_accuracy": 0.8364279270172119, "num_tokens": 91037840.0, "step": 2384 }, { "epoch": 0.3033965144383666, "ewc_loss": 0.01220703125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.11024284362793, "learning_rate": 1e-06, "loss": 0.4966, "mean_token_accuracy": 0.8391659259796143, "num_tokens": 91073336.0, "step": 2385 }, { "epoch": 0.30352372471695716, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.122371673583984, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8531860113143921, "num_tokens": 91107001.0, "step": 2386 }, { "epoch": 0.30365093499554763, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.114113807678223, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8591808080673218, "num_tokens": 91143937.0, "step": 2387 }, { "epoch": 0.30377814527413816, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.089271545410156, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8434913754463196, "num_tokens": 91185391.0, "step": 2388 }, { "epoch": 0.3039053555527287, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.098313331604004, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8511042594909668, "num_tokens": 91227653.0, "step": 2389 }, { "epoch": 0.30403256583131916, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.105755805969238, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8720699548721313, "num_tokens": 91263610.0, "step": 2390 }, { "epoch": 0.3041597761099097, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.123934745788574, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8473849296569824, "num_tokens": 91296880.0, "step": 2391 }, { "epoch": 0.3042869863885002, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.118249893188477, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8399962186813354, "num_tokens": 91336242.0, "step": 2392 }, { "epoch": 0.3044141966670907, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.103788375854492, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8601539134979248, "num_tokens": 91376527.0, "step": 2393 }, { "epoch": 0.3045414069456812, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.156829833984375, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8428353071212769, "num_tokens": 91408415.0, "step": 2394 }, { "epoch": 0.30466861722427174, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.068503379821777, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8534951210021973, "num_tokens": 91449189.0, "step": 2395 }, { "epoch": 0.3047958275028622, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.144104957580566, "learning_rate": 1e-06, "loss": 0.4693, "mean_token_accuracy": 0.8488452434539795, "num_tokens": 91488599.0, "step": 2396 }, { "epoch": 0.30492303778145274, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.102508544921875, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.856469988822937, "num_tokens": 91525138.0, "step": 2397 }, { "epoch": 0.30505024806004327, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.129042625427246, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8543930649757385, "num_tokens": 91567444.0, "step": 2398 }, { "epoch": 0.30517745833863374, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.155442237854004, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8493599891662598, "num_tokens": 91602926.0, "step": 2399 }, { "epoch": 0.3053046686172243, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.163671493530273, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8639485239982605, "num_tokens": 91635901.0, "step": 2400 }, { "epoch": 0.3054318788958148, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 9.079988479614258, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8548421859741211, "num_tokens": 91679542.0, "step": 2401 }, { "epoch": 0.3055590891744053, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.142062187194824, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8473240733146667, "num_tokens": 91723022.0, "step": 2402 }, { "epoch": 0.3056862994529958, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.228660583496094, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8490002155303955, "num_tokens": 91760664.0, "step": 2403 }, { "epoch": 0.30581350973158633, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.128247261047363, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8484464287757874, "num_tokens": 91805570.0, "step": 2404 }, { "epoch": 0.3059407200101768, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.262616157531738, "learning_rate": 1e-06, "loss": 0.5011, "mean_token_accuracy": 0.8438742160797119, "num_tokens": 91843342.0, "step": 2405 }, { "epoch": 0.30606793028876733, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.180196762084961, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.842981219291687, "num_tokens": 91883872.0, "step": 2406 }, { "epoch": 0.30619514056735786, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.138981819152832, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8511450290679932, "num_tokens": 91923174.0, "step": 2407 }, { "epoch": 0.30632235084594833, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.16633415222168, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8507674336433411, "num_tokens": 91963462.0, "step": 2408 }, { "epoch": 0.30644956112453886, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.14228630065918, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8558926582336426, "num_tokens": 92000907.0, "step": 2409 }, { "epoch": 0.3065767714031294, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.139788627624512, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8539891242980957, "num_tokens": 92036110.0, "step": 2410 }, { "epoch": 0.30670398168171986, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.15816879272461, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.84782874584198, "num_tokens": 92080647.0, "step": 2411 }, { "epoch": 0.3068311919603104, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.17679500579834, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8559457659721375, "num_tokens": 92115407.0, "step": 2412 }, { "epoch": 0.3069584022389009, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3484230041503906e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.133283615112305, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8525919914245605, "num_tokens": 92149660.0, "step": 2413 }, { "epoch": 0.3070856125174914, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.172873497009277, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8549345135688782, "num_tokens": 92191581.0, "step": 2414 }, { "epoch": 0.3072128227960819, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 9.138066291809082, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8471056818962097, "num_tokens": 92230435.0, "step": 2415 }, { "epoch": 0.30734003307467245, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.135490417480469, "learning_rate": 1e-06, "loss": 0.4984, "mean_token_accuracy": 0.8400297164916992, "num_tokens": 92269545.0, "step": 2416 }, { "epoch": 0.3074672433532629, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.191268920898438, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8588526248931885, "num_tokens": 92307447.0, "step": 2417 }, { "epoch": 0.30759445363185345, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.138537406921387, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8511311411857605, "num_tokens": 92349441.0, "step": 2418 }, { "epoch": 0.307721663910444, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.12732982635498, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8576769828796387, "num_tokens": 92380405.0, "step": 2419 }, { "epoch": 0.30784887418903445, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.165817260742188, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8500330448150635, "num_tokens": 92419596.0, "step": 2420 }, { "epoch": 0.307976084467625, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.205361366271973, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8456780910491943, "num_tokens": 92454590.0, "step": 2421 }, { "epoch": 0.3081032947462155, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.116430282592773, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8475475311279297, "num_tokens": 92491640.0, "step": 2422 }, { "epoch": 0.308230505024806, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.150486946105957, "learning_rate": 1e-06, "loss": 0.4926, "mean_token_accuracy": 0.8437225222587585, "num_tokens": 92528682.0, "step": 2423 }, { "epoch": 0.3083577153033965, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.153263092041016, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8586721420288086, "num_tokens": 92566368.0, "step": 2424 }, { "epoch": 0.30848492558198704, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.186861038208008, "learning_rate": 1e-06, "loss": 0.5213, "mean_token_accuracy": 0.8301459550857544, "num_tokens": 92601380.0, "step": 2425 }, { "epoch": 0.3086121358605775, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 9.107987403869629, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8677481412887573, "num_tokens": 92639724.0, "step": 2426 }, { "epoch": 0.30873934613916804, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.136421203613281, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8584297895431519, "num_tokens": 92677450.0, "step": 2427 }, { "epoch": 0.30886655641775856, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.12846851348877, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8608079552650452, "num_tokens": 92714681.0, "step": 2428 }, { "epoch": 0.3089937666963491, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.118590354919434, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8489525318145752, "num_tokens": 92749925.0, "step": 2429 }, { "epoch": 0.30912097697493957, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.151822090148926, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8566875457763672, "num_tokens": 92791720.0, "step": 2430 }, { "epoch": 0.3092481872535301, "ewc_loss": 0.012451171875, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.138669967651367, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8442345857620239, "num_tokens": 92826179.0, "step": 2431 }, { "epoch": 0.3093753975321206, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.114076614379883, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8513834476470947, "num_tokens": 92872751.0, "step": 2432 }, { "epoch": 0.3095026078107111, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.145722389221191, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8471838235855103, "num_tokens": 92907470.0, "step": 2433 }, { "epoch": 0.3096298180893016, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.179449081420898, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8572628498077393, "num_tokens": 92943321.0, "step": 2434 }, { "epoch": 0.30975702836789215, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.160082817077637, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8538954257965088, "num_tokens": 92982136.0, "step": 2435 }, { "epoch": 0.3098842386464826, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.197097778320312, "learning_rate": 1e-06, "loss": 0.4988, "mean_token_accuracy": 0.8419547080993652, "num_tokens": 93015974.0, "step": 2436 }, { "epoch": 0.31001144892507315, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.159881591796875, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8554698824882507, "num_tokens": 93054010.0, "step": 2437 }, { "epoch": 0.3101386592036637, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.071904182434082, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8601489067077637, "num_tokens": 93090226.0, "step": 2438 }, { "epoch": 0.31026586948225415, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.175226211547852, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8488468527793884, "num_tokens": 93130242.0, "step": 2439 }, { "epoch": 0.3103930797608447, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.146867752075195, "learning_rate": 1e-06, "loss": 0.4823, "mean_token_accuracy": 0.8482683897018433, "num_tokens": 93170990.0, "step": 2440 }, { "epoch": 0.3105202900394352, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.17022705078125, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8447530269622803, "num_tokens": 93205478.0, "step": 2441 }, { "epoch": 0.3106475003180257, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.125650405883789, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.8439252972602844, "num_tokens": 93240952.0, "step": 2442 }, { "epoch": 0.3107747105966162, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.171319007873535, "learning_rate": 1e-06, "loss": 0.4911, "mean_token_accuracy": 0.8417931795120239, "num_tokens": 93278882.0, "step": 2443 }, { "epoch": 0.31090192087520674, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.139947891235352, "learning_rate": 1e-06, "loss": 0.4832, "mean_token_accuracy": 0.8443417549133301, "num_tokens": 93324573.0, "step": 2444 }, { "epoch": 0.3110291311537972, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.123443603515625, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8661177158355713, "num_tokens": 93361049.0, "step": 2445 }, { "epoch": 0.31115634143238774, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.164632797241211, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8486990928649902, "num_tokens": 93401748.0, "step": 2446 }, { "epoch": 0.31128355171097827, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.180776596069336, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.863705039024353, "num_tokens": 93434857.0, "step": 2447 }, { "epoch": 0.31141076198956874, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.179840087890625, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8507510423660278, "num_tokens": 93474681.0, "step": 2448 }, { "epoch": 0.31153797226815927, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.136096954345703, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8504278659820557, "num_tokens": 93510627.0, "step": 2449 }, { "epoch": 0.3116651825467498, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.202672004699707, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8422353267669678, "num_tokens": 93551598.0, "step": 2450 }, { "epoch": 0.31179239282534027, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.155119895935059, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8446964621543884, "num_tokens": 93586622.0, "step": 2451 }, { "epoch": 0.3119196031039308, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.138349533081055, "learning_rate": 1e-06, "loss": 0.5067, "mean_token_accuracy": 0.8408989906311035, "num_tokens": 93626496.0, "step": 2452 }, { "epoch": 0.3120468133825213, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.200008392333984, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.853063702583313, "num_tokens": 93665641.0, "step": 2453 }, { "epoch": 0.3121740236611118, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.185860633850098, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8571051955223083, "num_tokens": 93702135.0, "step": 2454 }, { "epoch": 0.3123012339397023, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.196922302246094, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8586599230766296, "num_tokens": 93738226.0, "step": 2455 }, { "epoch": 0.31242844421829286, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.1612548828125, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8546237945556641, "num_tokens": 93780273.0, "step": 2456 }, { "epoch": 0.31255565449688333, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.155207633972168, "learning_rate": 1e-06, "loss": 0.4962, "mean_token_accuracy": 0.8391214609146118, "num_tokens": 93821977.0, "step": 2457 }, { "epoch": 0.31268286477547386, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.21297836303711, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8604156970977783, "num_tokens": 93856084.0, "step": 2458 }, { "epoch": 0.3128100750540644, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.220966339111328, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8555467128753662, "num_tokens": 93893491.0, "step": 2459 }, { "epoch": 0.31293728533265486, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.14041519165039, "learning_rate": 1e-06, "loss": 0.4924, "mean_token_accuracy": 0.8397071361541748, "num_tokens": 93940240.0, "step": 2460 }, { "epoch": 0.3130644956112454, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.206586837768555, "learning_rate": 1e-06, "loss": 0.4804, "mean_token_accuracy": 0.844628095626831, "num_tokens": 93974504.0, "step": 2461 }, { "epoch": 0.3131917058898359, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.189711570739746, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8533748984336853, "num_tokens": 94009208.0, "step": 2462 }, { "epoch": 0.3133189161684264, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 9.172460556030273, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8518688678741455, "num_tokens": 94049249.0, "step": 2463 }, { "epoch": 0.3134461264470169, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.178669929504395, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8527551293373108, "num_tokens": 94085553.0, "step": 2464 }, { "epoch": 0.31357333672560744, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.232231140136719, "learning_rate": 1e-06, "loss": 0.5134, "mean_token_accuracy": 0.8321059942245483, "num_tokens": 94127531.0, "step": 2465 }, { "epoch": 0.3137005470041979, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.116114616394043, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8636524677276611, "num_tokens": 94171191.0, "step": 2466 }, { "epoch": 0.31382775728278844, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.188471794128418, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8567652702331543, "num_tokens": 94212090.0, "step": 2467 }, { "epoch": 0.313954967561379, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.206863403320312, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8697267174720764, "num_tokens": 94246835.0, "step": 2468 }, { "epoch": 0.31408217783996945, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.163287162780762, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8503648042678833, "num_tokens": 94284581.0, "step": 2469 }, { "epoch": 0.31420938811856, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.207792282104492, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8615086078643799, "num_tokens": 94323050.0, "step": 2470 }, { "epoch": 0.3143365983971505, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.131753921508789, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.8471370935440063, "num_tokens": 94364702.0, "step": 2471 }, { "epoch": 0.314463808675741, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.21627426147461, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8583383560180664, "num_tokens": 94404735.0, "step": 2472 }, { "epoch": 0.3145910189543315, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.222893714904785, "learning_rate": 1e-06, "loss": 0.5012, "mean_token_accuracy": 0.8365521430969238, "num_tokens": 94444683.0, "step": 2473 }, { "epoch": 0.31471822923292203, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.15238094329834, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8542467355728149, "num_tokens": 94484392.0, "step": 2474 }, { "epoch": 0.3148454395115125, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.219797134399414, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8590930700302124, "num_tokens": 94524860.0, "step": 2475 }, { "epoch": 0.31497264979010303, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.169251441955566, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.8486241102218628, "num_tokens": 94565076.0, "step": 2476 }, { "epoch": 0.31509986006869356, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.175917625427246, "learning_rate": 1e-06, "loss": 0.5315, "mean_token_accuracy": 0.8293207883834839, "num_tokens": 94606191.0, "step": 2477 }, { "epoch": 0.31522707034728403, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.177818298339844, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.846331775188446, "num_tokens": 94648021.0, "step": 2478 }, { "epoch": 0.31535428062587456, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.20443344116211, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8628262877464294, "num_tokens": 94686305.0, "step": 2479 }, { "epoch": 0.3154814909044651, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.190428733825684, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8498015403747559, "num_tokens": 94728606.0, "step": 2480 }, { "epoch": 0.3156087011830556, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.178377151489258, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8547602891921997, "num_tokens": 94770162.0, "step": 2481 }, { "epoch": 0.3157359114616461, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.17971420288086, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8583270311355591, "num_tokens": 94807735.0, "step": 2482 }, { "epoch": 0.3158631217402366, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.245296478271484, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8463701605796814, "num_tokens": 94840366.0, "step": 2483 }, { "epoch": 0.31599033201882715, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.21230411529541, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8487343192100525, "num_tokens": 94874293.0, "step": 2484 }, { "epoch": 0.3161175422974176, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.185908317565918, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.851003885269165, "num_tokens": 94909347.0, "step": 2485 }, { "epoch": 0.31624475257600815, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.190905570983887, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8621920943260193, "num_tokens": 94955120.0, "step": 2486 }, { "epoch": 0.3163719628545987, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.233672142028809, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8662533760070801, "num_tokens": 94992319.0, "step": 2487 }, { "epoch": 0.31649917313318915, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.161958694458008, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8618417978286743, "num_tokens": 95023464.0, "step": 2488 }, { "epoch": 0.3166263834117797, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 9.23207950592041, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8651379942893982, "num_tokens": 95055963.0, "step": 2489 }, { "epoch": 0.3167535936903702, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.233166694641113, "learning_rate": 1e-06, "loss": 0.513, "mean_token_accuracy": 0.8341642618179321, "num_tokens": 95089628.0, "step": 2490 }, { "epoch": 0.3168808039689607, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 9.15543270111084, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8659064769744873, "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": 9.22785758972168, "learning_rate": 1e-06, "loss": 0.4941, "mean_token_accuracy": 0.8402711153030396, "num_tokens": 95168324.0, "step": 2492 }, { "epoch": 0.31713522452614173, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.220349311828613, "learning_rate": 1e-06, "loss": 0.5268, "mean_token_accuracy": 0.8327487707138062, "num_tokens": 95208139.0, "step": 2493 }, { "epoch": 0.3172624348047322, "ewc_loss": 0.0126953125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 9.162878036499023, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8467479944229126, "num_tokens": 95247337.0, "step": 2494 }, { "epoch": 0.31738964508332274, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.171195983886719, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8450905084609985, "num_tokens": 95289208.0, "step": 2495 }, { "epoch": 0.31751685536191326, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.164266586303711, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8504669070243835, "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": 9.148523330688477, "learning_rate": 1e-06, "loss": 0.5066, "mean_token_accuracy": 0.8316982388496399, "num_tokens": 95371052.0, "step": 2497 }, { "epoch": 0.31777127591909426, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.123446464538574, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8717001080513, "num_tokens": 95412486.0, "step": 2498 }, { "epoch": 0.3178984861976848, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.207392692565918, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8681904077529907, "num_tokens": 95446167.0, "step": 2499 }, { "epoch": 0.31802569647627527, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.172614097595215, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8468546867370605, "num_tokens": 95485763.0, "step": 2500 }, { "epoch": 0.3181529067548658, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.179320335388184, "learning_rate": 1e-06, "loss": 0.5206, "mean_token_accuracy": 0.8416812419891357, "num_tokens": 95526407.0, "step": 2501 }, { "epoch": 0.3182801170334563, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.222522735595703, "learning_rate": 1e-06, "loss": 0.5626, "mean_token_accuracy": 0.8200005292892456, "num_tokens": 95567479.0, "step": 2502 }, { "epoch": 0.3184073273120468, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.135053634643555, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8586299419403076, "num_tokens": 95606588.0, "step": 2503 }, { "epoch": 0.3185345375906373, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 9.186527252197266, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8561766147613525, "num_tokens": 95653293.0, "step": 2504 }, { "epoch": 0.31866174786922785, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.217491149902344, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.8434810638427734, "num_tokens": 95692424.0, "step": 2505 }, { "epoch": 0.3187889581478183, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.152459144592285, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8614461421966553, "num_tokens": 95733679.0, "step": 2506 }, { "epoch": 0.31891616842640885, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.243078231811523, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8474156856536865, "num_tokens": 95768211.0, "step": 2507 }, { "epoch": 0.3190433787049994, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.217924118041992, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8508988618850708, "num_tokens": 95805974.0, "step": 2508 }, { "epoch": 0.31917058898358985, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.23483657836914, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.850845456123352, "num_tokens": 95845463.0, "step": 2509 }, { "epoch": 0.3192977992621804, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.19303035736084, "learning_rate": 1e-06, "loss": 0.5084, "mean_token_accuracy": 0.847576916217804, "num_tokens": 95888148.0, "step": 2510 }, { "epoch": 0.3194250095407709, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.21741008758545, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8732419013977051, "num_tokens": 95925432.0, "step": 2511 }, { "epoch": 0.3195522198193614, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.169593811035156, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8580613732337952, "num_tokens": 95970695.0, "step": 2512 }, { "epoch": 0.3196794300979519, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.204771041870117, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8525103330612183, "num_tokens": 96006093.0, "step": 2513 }, { "epoch": 0.31980664037654244, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.21543025970459, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.851830244064331, "num_tokens": 96045945.0, "step": 2514 }, { "epoch": 0.3199338506551329, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.250516891479492, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8687120079994202, "num_tokens": 96076748.0, "step": 2515 }, { "epoch": 0.32006106093372344, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.190999031066895, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8538999557495117, "num_tokens": 96116964.0, "step": 2516 }, { "epoch": 0.32018827121231397, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.24627685546875, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8547866344451904, "num_tokens": 96152073.0, "step": 2517 }, { "epoch": 0.32031548149090444, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 9.213907241821289, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8510221242904663, "num_tokens": 96191608.0, "step": 2518 }, { "epoch": 0.32044269176949497, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.205888748168945, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8701371550559998, "num_tokens": 96228124.0, "step": 2519 }, { "epoch": 0.3205699020480855, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.18008804321289, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8631355166435242, "num_tokens": 96265870.0, "step": 2520 }, { "epoch": 0.32069711232667597, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.20801067352295, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8624457716941833, "num_tokens": 96309623.0, "step": 2521 }, { "epoch": 0.3208243226052665, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.21088981628418, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8450616598129272, "num_tokens": 96344036.0, "step": 2522 }, { "epoch": 0.320951532883857, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.275863647460938, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8634037971496582, "num_tokens": 96377627.0, "step": 2523 }, { "epoch": 0.3210787431624475, "ewc_loss": 0.012939453125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 9.186200141906738, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8542847633361816, "num_tokens": 96413568.0, "step": 2524 }, { "epoch": 0.32120595344103803, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.22447395324707, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8504235744476318, "num_tokens": 96454682.0, "step": 2525 }, { "epoch": 0.32133316371962856, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 9.178746223449707, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.841774046421051, "num_tokens": 96490456.0, "step": 2526 }, { "epoch": 0.32146037399821903, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.272149085998535, "learning_rate": 1e-06, "loss": 0.5019, "mean_token_accuracy": 0.8458890318870544, "num_tokens": 96526891.0, "step": 2527 }, { "epoch": 0.32158758427680956, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.172904014587402, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8471454977989197, "num_tokens": 96575226.0, "step": 2528 }, { "epoch": 0.3217147945554001, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.264720916748047, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8422375321388245, "num_tokens": 96606383.0, "step": 2529 }, { "epoch": 0.3218420048339906, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.190237045288086, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8551071882247925, "num_tokens": 96639826.0, "step": 2530 }, { "epoch": 0.3219692151125811, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.168573379516602, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.856725811958313, "num_tokens": 96679151.0, "step": 2531 }, { "epoch": 0.3220964253911716, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.283499717712402, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.859164834022522, "num_tokens": 96719365.0, "step": 2532 }, { "epoch": 0.32222363566976214, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.239690780639648, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8511484861373901, "num_tokens": 96758318.0, "step": 2533 }, { "epoch": 0.3223508459483526, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.244967460632324, "learning_rate": 1e-06, "loss": 0.4889, "mean_token_accuracy": 0.8477532863616943, "num_tokens": 96791159.0, "step": 2534 }, { "epoch": 0.32247805622694314, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.190945625305176, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8556514978408813, "num_tokens": 96825352.0, "step": 2535 }, { "epoch": 0.32260526650553367, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.253609657287598, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8508487939834595, "num_tokens": 96859700.0, "step": 2536 }, { "epoch": 0.32273247678412414, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.218894004821777, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8502211570739746, "num_tokens": 96897077.0, "step": 2537 }, { "epoch": 0.3228596870627147, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.184497833251953, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8471196293830872, "num_tokens": 96936936.0, "step": 2538 }, { "epoch": 0.3229868973413052, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.21582317352295, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8580456972122192, "num_tokens": 96975702.0, "step": 2539 }, { "epoch": 0.3231141076198957, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.220240592956543, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8513289093971252, "num_tokens": 97010585.0, "step": 2540 }, { "epoch": 0.3232413178984862, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.194302558898926, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8762015104293823, "num_tokens": 97047210.0, "step": 2541 }, { "epoch": 0.32336852817707673, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.231242179870605, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8472713828086853, "num_tokens": 97084835.0, "step": 2542 }, { "epoch": 0.3234957384556672, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.1975679397583, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8505301475524902, "num_tokens": 97125829.0, "step": 2543 }, { "epoch": 0.32362294873425773, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.153523445129395, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8464035987854004, "num_tokens": 97162874.0, "step": 2544 }, { "epoch": 0.32375015901284826, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.264324188232422, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8558098077774048, "num_tokens": 97197559.0, "step": 2545 }, { "epoch": 0.32387736929143873, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.244494438171387, "learning_rate": 1e-06, "loss": 0.4945, "mean_token_accuracy": 0.8401397466659546, "num_tokens": 97227028.0, "step": 2546 }, { "epoch": 0.32400457957002926, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.19675350189209, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8483481407165527, "num_tokens": 97270209.0, "step": 2547 }, { "epoch": 0.3241317898486198, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.210055351257324, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8555601239204407, "num_tokens": 97309176.0, "step": 2548 }, { "epoch": 0.32425900012721026, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.261983871459961, "learning_rate": 1e-06, "loss": 0.512, "mean_token_accuracy": 0.833445131778717, "num_tokens": 97340464.0, "step": 2549 }, { "epoch": 0.3243862104058008, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.206254959106445, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8552825450897217, "num_tokens": 97378435.0, "step": 2550 }, { "epoch": 0.3245134206843913, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.250797271728516, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8485170602798462, "num_tokens": 97417886.0, "step": 2551 }, { "epoch": 0.3246406309629818, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.258807182312012, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.855643630027771, "num_tokens": 97450552.0, "step": 2552 }, { "epoch": 0.3247678412415723, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.252522468566895, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8595637679100037, "num_tokens": 97488586.0, "step": 2553 }, { "epoch": 0.32489505152016285, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.22014331817627, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8478811383247375, "num_tokens": 97528893.0, "step": 2554 }, { "epoch": 0.3250222617987533, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.279500961303711, "learning_rate": 1e-06, "loss": 0.5277, "mean_token_accuracy": 0.8289752006530762, "num_tokens": 97568654.0, "step": 2555 }, { "epoch": 0.32514947207734385, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.236928939819336, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8561741709709167, "num_tokens": 97607068.0, "step": 2556 }, { "epoch": 0.3252766823559344, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.257048606872559, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8742769360542297, "num_tokens": 97647505.0, "step": 2557 }, { "epoch": 0.32540389263452485, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.245078086853027, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8465370535850525, "num_tokens": 97690505.0, "step": 2558 }, { "epoch": 0.3255311029131154, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.28398609161377, "learning_rate": 1e-06, "loss": 0.5099, "mean_token_accuracy": 0.840427041053772, "num_tokens": 97723955.0, "step": 2559 }, { "epoch": 0.3256583131917059, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.29306411743164, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.846459150314331, "num_tokens": 97763822.0, "step": 2560 }, { "epoch": 0.3257855234702964, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.228901863098145, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8481935858726501, "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": 9.270657539367676, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8514078855514526, "num_tokens": 97841056.0, "step": 2562 }, { "epoch": 0.32603994402747744, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.243107795715332, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.869166910648346, "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": 9.277225494384766, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8500375747680664, "num_tokens": 97925797.0, "step": 2564 }, { "epoch": 0.32629436458465844, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.331128120422363, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8568945527076721, "num_tokens": 97963638.0, "step": 2565 }, { "epoch": 0.32642157486324896, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.249421119689941, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8465470671653748, "num_tokens": 98002043.0, "step": 2566 }, { "epoch": 0.32654878514183944, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.275472640991211, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.855161726474762, "num_tokens": 98041031.0, "step": 2567 }, { "epoch": 0.32667599542042997, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.312311172485352, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8599627614021301, "num_tokens": 98075510.0, "step": 2568 }, { "epoch": 0.3268032056990205, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.228728294372559, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8587864637374878, "num_tokens": 98114162.0, "step": 2569 }, { "epoch": 0.32693041597761097, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.361743927001953, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8479096293449402, "num_tokens": 98148722.0, "step": 2570 }, { "epoch": 0.3270576262562015, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.263482093811035, "learning_rate": 1e-06, "loss": 0.5278, "mean_token_accuracy": 0.8319745063781738, "num_tokens": 98188255.0, "step": 2571 }, { "epoch": 0.327184836534792, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.26812744140625, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8634070754051208, "num_tokens": 98228597.0, "step": 2572 }, { "epoch": 0.3273120468133825, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.26548957824707, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8434631824493408, "num_tokens": 98273405.0, "step": 2573 }, { "epoch": 0.327439257091973, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.261839866638184, "learning_rate": 1e-06, "loss": 0.5178, "mean_token_accuracy": 0.8382583856582642, "num_tokens": 98315091.0, "step": 2574 }, { "epoch": 0.32756646737056355, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.240830421447754, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.854154646396637, "num_tokens": 98356967.0, "step": 2575 }, { "epoch": 0.327693677649154, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.268806457519531, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8434301018714905, "num_tokens": 98394455.0, "step": 2576 }, { "epoch": 0.32782088792774455, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.222126007080078, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8609499931335449, "num_tokens": 98430455.0, "step": 2577 }, { "epoch": 0.3279480982063351, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 9.277438163757324, "learning_rate": 1e-06, "loss": 0.4974, "mean_token_accuracy": 0.8386321067810059, "num_tokens": 98472218.0, "step": 2578 }, { "epoch": 0.3280753084849256, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.280414581298828, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8596093654632568, "num_tokens": 98509540.0, "step": 2579 }, { "epoch": 0.3282025187635161, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.252082824707031, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.851325511932373, "num_tokens": 98544761.0, "step": 2580 }, { "epoch": 0.3283297290421066, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.318096160888672, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8651857376098633, "num_tokens": 98580522.0, "step": 2581 }, { "epoch": 0.32845693932069714, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.262666702270508, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8608711957931519, "num_tokens": 98612988.0, "step": 2582 }, { "epoch": 0.3285841495992876, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.261441230773926, "learning_rate": 1e-06, "loss": 0.4873, "mean_token_accuracy": 0.8402671813964844, "num_tokens": 98654213.0, "step": 2583 }, { "epoch": 0.32871135987787814, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.239968299865723, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8551052212715149, "num_tokens": 98697168.0, "step": 2584 }, { "epoch": 0.32883857015646867, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.277227401733398, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8573659062385559, "num_tokens": 98735664.0, "step": 2585 }, { "epoch": 0.32896578043505914, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.285440444946289, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8473953604698181, "num_tokens": 98777257.0, "step": 2586 }, { "epoch": 0.32909299071364967, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.319549560546875, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8568012714385986, "num_tokens": 98812443.0, "step": 2587 }, { "epoch": 0.3292202009922402, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.27481746673584, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8609380125999451, "num_tokens": 98848231.0, "step": 2588 }, { "epoch": 0.32934741127083067, "ewc_loss": 0.01318359375, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 9.281749725341797, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8557467460632324, "num_tokens": 98886487.0, "step": 2589 }, { "epoch": 0.3294746215494212, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.270333290100098, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8430424928665161, "num_tokens": 98924750.0, "step": 2590 }, { "epoch": 0.3296018318280117, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.230560302734375, "learning_rate": 1e-06, "loss": 0.5386, "mean_token_accuracy": 0.8299442529678345, "num_tokens": 98964151.0, "step": 2591 }, { "epoch": 0.3297290421066022, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.284782409667969, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8640789985656738, "num_tokens": 99000277.0, "step": 2592 }, { "epoch": 0.3298562523851927, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.254752159118652, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8428463935852051, "num_tokens": 99042335.0, "step": 2593 }, { "epoch": 0.32998346266378326, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.235751152038574, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8514194488525391, "num_tokens": 99082964.0, "step": 2594 }, { "epoch": 0.33011067294237373, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.264171600341797, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8761718273162842, "num_tokens": 99123440.0, "step": 2595 }, { "epoch": 0.33023788322096426, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.263232231140137, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8512349724769592, "num_tokens": 99167047.0, "step": 2596 }, { "epoch": 0.3303650934995548, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.270730972290039, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8499842882156372, "num_tokens": 99202594.0, "step": 2597 }, { "epoch": 0.33049230377814526, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.246685981750488, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8590748310089111, "num_tokens": 99244732.0, "step": 2598 }, { "epoch": 0.3306195140567358, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.285311698913574, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8630924224853516, "num_tokens": 99287161.0, "step": 2599 }, { "epoch": 0.3307467243353263, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.270523071289062, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8453332185745239, "num_tokens": 99322650.0, "step": 2600 }, { "epoch": 0.3308739346139168, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.352190017700195, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8532302379608154, "num_tokens": 99358529.0, "step": 2601 }, { "epoch": 0.3310011448925073, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.251360893249512, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8616454601287842, "num_tokens": 99394010.0, "step": 2602 }, { "epoch": 0.33112835517109784, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.312155723571777, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8556718230247498, "num_tokens": 99434764.0, "step": 2603 }, { "epoch": 0.3312555654496883, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.241204261779785, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.854735791683197, "num_tokens": 99481322.0, "step": 2604 }, { "epoch": 0.33138277572827884, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.296830177307129, "learning_rate": 1e-06, "loss": 0.4998, "mean_token_accuracy": 0.8437191843986511, "num_tokens": 99517719.0, "step": 2605 }, { "epoch": 0.3315099860068694, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.282464981079102, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8550065159797668, "num_tokens": 99553662.0, "step": 2606 }, { "epoch": 0.33163719628545985, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.302810668945312, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8526278138160706, "num_tokens": 99589195.0, "step": 2607 }, { "epoch": 0.3317644065640504, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.253106117248535, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8556087017059326, "num_tokens": 99626092.0, "step": 2608 }, { "epoch": 0.3318916168426409, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6226043701171875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.309874534606934, "learning_rate": 1e-06, "loss": 0.5256, "mean_token_accuracy": 0.8279998302459717, "num_tokens": 99667087.0, "step": 2609 }, { "epoch": 0.3320188271212314, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.259820938110352, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8477933406829834, "num_tokens": 99705929.0, "step": 2610 }, { "epoch": 0.3321460373998219, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.241345405578613, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8495911359786987, "num_tokens": 99742044.0, "step": 2611 }, { "epoch": 0.33227324767841243, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.321685791015625, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8556104302406311, "num_tokens": 99772621.0, "step": 2612 }, { "epoch": 0.3324004579570029, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.250167846679688, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8584842681884766, "num_tokens": 99814405.0, "step": 2613 }, { "epoch": 0.33252766823559343, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.2506685256958, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8634114861488342, "num_tokens": 99852197.0, "step": 2614 }, { "epoch": 0.33265487851418396, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.292231559753418, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8483787775039673, "num_tokens": 99894254.0, "step": 2615 }, { "epoch": 0.33278208879277443, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.249218940734863, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8474634885787964, "num_tokens": 99936230.0, "step": 2616 }, { "epoch": 0.33290929907136496, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.285898208618164, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8538511395454407, "num_tokens": 99974964.0, "step": 2617 }, { "epoch": 0.3330365093499555, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.273824691772461, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8436967730522156, "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": 9.314767837524414, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.8477689027786255, "num_tokens": 100054064.0, "step": 2619 }, { "epoch": 0.3332909299071365, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.33098316192627, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.857599139213562, "num_tokens": 100092549.0, "step": 2620 }, { "epoch": 0.333418140185727, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.298412322998047, "learning_rate": 1e-06, "loss": 0.5353, "mean_token_accuracy": 0.8274355530738831, "num_tokens": 100123973.0, "step": 2621 }, { "epoch": 0.3335453504643175, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 9.261852264404297, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8515481948852539, "num_tokens": 100162142.0, "step": 2622 }, { "epoch": 0.333672560742908, "ewc_loss": 0.013427734375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.243596076965332, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8477264046669006, "num_tokens": 100199352.0, "step": 2623 }, { "epoch": 0.33379977102149855, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.294504165649414, "learning_rate": 1e-06, "loss": 0.505, "mean_token_accuracy": 0.8366605043411255, "num_tokens": 100237446.0, "step": 2624 }, { "epoch": 0.333926981300089, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.237669944763184, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8575483560562134, "num_tokens": 100272986.0, "step": 2625 }, { "epoch": 0.33405419157867955, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 9.29615306854248, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8439851999282837, "num_tokens": 100304998.0, "step": 2626 }, { "epoch": 0.3341814018572701, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.2555513381958, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8484818935394287, "num_tokens": 100338473.0, "step": 2627 }, { "epoch": 0.33430861213586055, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.293732643127441, "learning_rate": 1e-06, "loss": 0.5202, "mean_token_accuracy": 0.836505651473999, "num_tokens": 100372644.0, "step": 2628 }, { "epoch": 0.3344358224144511, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.257575035095215, "learning_rate": 1e-06, "loss": 0.5214, "mean_token_accuracy": 0.8333329558372498, "num_tokens": 100413248.0, "step": 2629 }, { "epoch": 0.3345630326930416, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.282536506652832, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8468276262283325, "num_tokens": 100448022.0, "step": 2630 }, { "epoch": 0.33469024297163213, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.29643440246582, "learning_rate": 1e-06, "loss": 0.4977, "mean_token_accuracy": 0.8407542705535889, "num_tokens": 100493222.0, "step": 2631 }, { "epoch": 0.3348174532502226, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.26242733001709, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.858238160610199, "num_tokens": 100534736.0, "step": 2632 }, { "epoch": 0.33494466352881314, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.297957420349121, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8593050241470337, "num_tokens": 100576754.0, "step": 2633 }, { "epoch": 0.33507187380740366, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.30809211730957, "learning_rate": 1e-06, "loss": 0.4816, "mean_token_accuracy": 0.844315230846405, "num_tokens": 100611407.0, "step": 2634 }, { "epoch": 0.33519908408599414, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.294845581054688, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8589425086975098, "num_tokens": 100643531.0, "step": 2635 }, { "epoch": 0.33532629436458466, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.280597686767578, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8530066013336182, "num_tokens": 100678612.0, "step": 2636 }, { "epoch": 0.3354535046431752, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.288982391357422, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8566941022872925, "num_tokens": 100718873.0, "step": 2637 }, { "epoch": 0.33558071492176567, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.305559158325195, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8511912822723389, "num_tokens": 100759883.0, "step": 2638 }, { "epoch": 0.3357079252003562, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.27867603302002, "learning_rate": 1e-06, "loss": 0.5162, "mean_token_accuracy": 0.8325452208518982, "num_tokens": 100800022.0, "step": 2639 }, { "epoch": 0.3358351354789467, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.331053733825684, "learning_rate": 1e-06, "loss": 0.5113, "mean_token_accuracy": 0.8372465372085571, "num_tokens": 100837299.0, "step": 2640 }, { "epoch": 0.3359623457575372, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.297706604003906, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.849082350730896, "num_tokens": 100868433.0, "step": 2641 }, { "epoch": 0.3360895560361277, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.328969955444336, "learning_rate": 1e-06, "loss": 0.4971, "mean_token_accuracy": 0.8373676538467407, "num_tokens": 100907768.0, "step": 2642 }, { "epoch": 0.33621676631471825, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.274961471557617, "learning_rate": 1e-06, "loss": 0.5301, "mean_token_accuracy": 0.8295634984970093, "num_tokens": 100942723.0, "step": 2643 }, { "epoch": 0.3363439765933087, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.307717323303223, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8573793172836304, "num_tokens": 100984997.0, "step": 2644 }, { "epoch": 0.33647118687189925, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.235003471374512, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8665652871131897, "num_tokens": 101026290.0, "step": 2645 }, { "epoch": 0.3365983971504898, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.403937339782715, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.8450469374656677, "num_tokens": 101066766.0, "step": 2646 }, { "epoch": 0.33672560742908025, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.310861587524414, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8557671904563904, "num_tokens": 101100833.0, "step": 2647 }, { "epoch": 0.3368528177076708, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.32534408569336, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8621205687522888, "num_tokens": 101145552.0, "step": 2648 }, { "epoch": 0.3369800279862613, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.261919975280762, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8541706800460815, "num_tokens": 101186644.0, "step": 2649 }, { "epoch": 0.3371072382648518, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 9.343454360961914, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8495720624923706, "num_tokens": 101223356.0, "step": 2650 }, { "epoch": 0.3372344485434423, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.315982818603516, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8642680644989014, "num_tokens": 101260527.0, "step": 2651 }, { "epoch": 0.33736165882203284, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.231962203979492, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8516825437545776, "num_tokens": 101305403.0, "step": 2652 }, { "epoch": 0.3374888691006233, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 9.291361808776855, "learning_rate": 1e-06, "loss": 0.5135, "mean_token_accuracy": 0.8374103307723999, "num_tokens": 101345858.0, "step": 2653 }, { "epoch": 0.33761607937921384, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.342095375061035, "learning_rate": 1e-06, "loss": 0.5254, "mean_token_accuracy": 0.8286436796188354, "num_tokens": 101380321.0, "step": 2654 }, { "epoch": 0.33774328965780437, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.273807525634766, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8613516092300415, "num_tokens": 101419120.0, "step": 2655 }, { "epoch": 0.33787049993639484, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.365203857421875, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8543922901153564, "num_tokens": 101453272.0, "step": 2656 }, { "epoch": 0.33799771021498537, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.362679481506348, "learning_rate": 1e-06, "loss": 0.4997, "mean_token_accuracy": 0.8415483236312866, "num_tokens": 101486879.0, "step": 2657 }, { "epoch": 0.3381249204935759, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.330521583557129, "learning_rate": 1e-06, "loss": 0.4825, "mean_token_accuracy": 0.8443148732185364, "num_tokens": 101523691.0, "step": 2658 }, { "epoch": 0.33825213077216637, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.281360626220703, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8533723950386047, "num_tokens": 101555279.0, "step": 2659 }, { "epoch": 0.3383793410507569, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.275951385498047, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8683109283447266, "num_tokens": 101591780.0, "step": 2660 }, { "epoch": 0.3385065513293474, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.275689125061035, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.86765056848526, "num_tokens": 101631041.0, "step": 2661 }, { "epoch": 0.3386337616079379, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.295785903930664, "learning_rate": 1e-06, "loss": 0.4902, "mean_token_accuracy": 0.8416568040847778, "num_tokens": 101666315.0, "step": 2662 }, { "epoch": 0.33876097188652843, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.296026229858398, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8651509284973145, "num_tokens": 101704046.0, "step": 2663 }, { "epoch": 0.33888818216511896, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.269933700561523, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8628774881362915, "num_tokens": 101743076.0, "step": 2664 }, { "epoch": 0.33901539244370943, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.295771598815918, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8656452894210815, "num_tokens": 101784006.0, "step": 2665 }, { "epoch": 0.33914260272229996, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.321396827697754, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.859045147895813, "num_tokens": 101821845.0, "step": 2666 }, { "epoch": 0.3392698130008905, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.253182411193848, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.857698917388916, "num_tokens": 101861121.0, "step": 2667 }, { "epoch": 0.33939702327948096, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.3284330368042, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8490561246871948, "num_tokens": 101894330.0, "step": 2668 }, { "epoch": 0.3395242335580715, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.319097518920898, "learning_rate": 1e-06, "loss": 0.5043, "mean_token_accuracy": 0.8371522426605225, "num_tokens": 101934835.0, "step": 2669 }, { "epoch": 0.339651443836662, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.299773216247559, "learning_rate": 1e-06, "loss": 0.513, "mean_token_accuracy": 0.8355271816253662, "num_tokens": 101970943.0, "step": 2670 }, { "epoch": 0.3397786541152525, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 9.332730293273926, "learning_rate": 1e-06, "loss": 0.4897, "mean_token_accuracy": 0.8403435945510864, "num_tokens": 102001211.0, "step": 2671 }, { "epoch": 0.339905864393843, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.30447006225586, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.858923077583313, "num_tokens": 102039399.0, "step": 2672 }, { "epoch": 0.34003307467243354, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.32003116607666, "learning_rate": 1e-06, "loss": 0.4909, "mean_token_accuracy": 0.8418906331062317, "num_tokens": 102070635.0, "step": 2673 }, { "epoch": 0.340160284951024, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.327803611755371, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8568477630615234, "num_tokens": 102104590.0, "step": 2674 }, { "epoch": 0.34028749522961454, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.27830696105957, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8583292365074158, "num_tokens": 102149267.0, "step": 2675 }, { "epoch": 0.3404147055082051, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7179718017578125e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.317294120788574, "learning_rate": 1e-06, "loss": 0.5245, "mean_token_accuracy": 0.8365464210510254, "num_tokens": 102188089.0, "step": 2676 }, { "epoch": 0.34054191578679555, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.311859130859375, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8593353629112244, "num_tokens": 102224335.0, "step": 2677 }, { "epoch": 0.3406691260653861, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.329605102539062, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8562612533569336, "num_tokens": 102263421.0, "step": 2678 }, { "epoch": 0.3407963363439766, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.322830200195312, "learning_rate": 1e-06, "loss": 0.5195, "mean_token_accuracy": 0.8356343507766724, "num_tokens": 102301133.0, "step": 2679 }, { "epoch": 0.34092354662256713, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.349265098571777, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8569613695144653, "num_tokens": 102338516.0, "step": 2680 }, { "epoch": 0.3410507569011576, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.327515602111816, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8432061076164246, "num_tokens": 102376010.0, "step": 2681 }, { "epoch": 0.34117796717974813, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.337037086486816, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8734192848205566, "num_tokens": 102417080.0, "step": 2682 }, { "epoch": 0.34130517745833866, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.321642875671387, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.855053186416626, "num_tokens": 102458042.0, "step": 2683 }, { "epoch": 0.34143238773692913, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.312246322631836, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8610526919364929, "num_tokens": 102497152.0, "step": 2684 }, { "epoch": 0.34155959801551966, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.355449676513672, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8507156372070312, "num_tokens": 102531486.0, "step": 2685 }, { "epoch": 0.3416868082941102, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.31531047821045, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8516674637794495, "num_tokens": 102571552.0, "step": 2686 }, { "epoch": 0.34181401857270066, "ewc_loss": 0.01373291015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 9.319469451904297, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8588864803314209, "num_tokens": 102607338.0, "step": 2687 }, { "epoch": 0.3419412288512912, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.359808921813965, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8529619574546814, "num_tokens": 102654445.0, "step": 2688 }, { "epoch": 0.3420684391298817, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.284679412841797, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8616938591003418, "num_tokens": 102695265.0, "step": 2689 }, { "epoch": 0.3421956494084722, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.276368141174316, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8572750091552734, "num_tokens": 102728621.0, "step": 2690 }, { "epoch": 0.3423228596870627, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.345540046691895, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8467885255813599, "num_tokens": 102766378.0, "step": 2691 }, { "epoch": 0.34245006996565325, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 9.296627044677734, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8518387079238892, "num_tokens": 102806331.0, "step": 2692 }, { "epoch": 0.3425772802442437, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.324056625366211, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8562105298042297, "num_tokens": 102845314.0, "step": 2693 }, { "epoch": 0.34270449052283425, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.345552444458008, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8517453670501709, "num_tokens": 102888614.0, "step": 2694 }, { "epoch": 0.3428317008014248, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.309502601623535, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8480843901634216, "num_tokens": 102925988.0, "step": 2695 }, { "epoch": 0.34295891108001525, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.315545082092285, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8611936569213867, "num_tokens": 102963150.0, "step": 2696 }, { "epoch": 0.3430861213586058, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.364072799682617, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8609377145767212, "num_tokens": 102999164.0, "step": 2697 }, { "epoch": 0.3432133316371963, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.325479507446289, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.861648440361023, "num_tokens": 103034337.0, "step": 2698 }, { "epoch": 0.3433405419157868, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.36343765258789, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8592992424964905, "num_tokens": 103075062.0, "step": 2699 }, { "epoch": 0.3434677521943773, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.350675582885742, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8483688235282898, "num_tokens": 103113209.0, "step": 2700 }, { "epoch": 0.34359496247296784, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.379578590393066, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8573076128959656, "num_tokens": 103153117.0, "step": 2701 }, { "epoch": 0.3437221727515583, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.286847114562988, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8518097400665283, "num_tokens": 103192234.0, "step": 2702 }, { "epoch": 0.34384938303014884, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.347724914550781, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8498088717460632, "num_tokens": 103224958.0, "step": 2703 }, { "epoch": 0.34397659330873936, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.355095863342285, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8554003834724426, "num_tokens": 103259679.0, "step": 2704 }, { "epoch": 0.34410380358732984, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.328676223754883, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8443847894668579, "num_tokens": 103299987.0, "step": 2705 }, { "epoch": 0.34423101386592037, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.35914134979248, "learning_rate": 1e-06, "loss": 0.4989, "mean_token_accuracy": 0.8394439816474915, "num_tokens": 103344457.0, "step": 2706 }, { "epoch": 0.3443582241445109, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.27521800994873, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8589056134223938, "num_tokens": 103385173.0, "step": 2707 }, { "epoch": 0.34448543442310137, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.351517677307129, "learning_rate": 1e-06, "loss": 0.5048, "mean_token_accuracy": 0.8384513854980469, "num_tokens": 103425972.0, "step": 2708 }, { "epoch": 0.3446126447016919, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.35557746887207, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8494338393211365, "num_tokens": 103466819.0, "step": 2709 }, { "epoch": 0.3447398549802824, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.7418136596679688e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.308426856994629, "learning_rate": 1e-06, "loss": 0.4903, "mean_token_accuracy": 0.8419402837753296, "num_tokens": 103509660.0, "step": 2710 }, { "epoch": 0.3448670652588729, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.385679244995117, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8475332260131836, "num_tokens": 103544557.0, "step": 2711 }, { "epoch": 0.3449942755374634, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.320807456970215, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8624826669692993, "num_tokens": 103585132.0, "step": 2712 }, { "epoch": 0.34512148581605395, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.303013801574707, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8620539307594299, "num_tokens": 103626353.0, "step": 2713 }, { "epoch": 0.3452486960946444, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.365692138671875, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8565517067909241, "num_tokens": 103659035.0, "step": 2714 }, { "epoch": 0.34537590637323495, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.343639373779297, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.861038327217102, "num_tokens": 103699497.0, "step": 2715 }, { "epoch": 0.3455031166518255, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.355588912963867, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8414430618286133, "num_tokens": 103733685.0, "step": 2716 }, { "epoch": 0.34563032693041595, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.321722030639648, "learning_rate": 1e-06, "loss": 0.5054, "mean_token_accuracy": 0.8362311124801636, "num_tokens": 103769399.0, "step": 2717 }, { "epoch": 0.3457575372090065, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.319584846496582, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.860555112361908, "num_tokens": 103809078.0, "step": 2718 }, { "epoch": 0.345884747487597, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.342794418334961, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8699086308479309, "num_tokens": 103846764.0, "step": 2719 }, { "epoch": 0.3460119577661875, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.323861122131348, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8623404502868652, "num_tokens": 103883089.0, "step": 2720 }, { "epoch": 0.346139168044778, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.312177658081055, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8710322380065918, "num_tokens": 103915775.0, "step": 2721 }, { "epoch": 0.34626637832336854, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.372004508972168, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8483636975288391, "num_tokens": 103953703.0, "step": 2722 }, { "epoch": 0.346393588601959, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.349093437194824, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8433943390846252, "num_tokens": 103990775.0, "step": 2723 }, { "epoch": 0.34652079888054954, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.360007286071777, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8575770258903503, "num_tokens": 104025927.0, "step": 2724 }, { "epoch": 0.34664800915914007, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.340032577514648, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8442360758781433, "num_tokens": 104065860.0, "step": 2725 }, { "epoch": 0.34677521943773054, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.402239799499512, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8557502627372742, "num_tokens": 104102117.0, "step": 2726 }, { "epoch": 0.34690242971632107, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.318585395812988, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8498374223709106, "num_tokens": 104140896.0, "step": 2727 }, { "epoch": 0.3470296399949116, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.367606163024902, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.851926326751709, "num_tokens": 104181355.0, "step": 2728 }, { "epoch": 0.3471568502735021, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.34676742553711, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8547873497009277, "num_tokens": 104218564.0, "step": 2729 }, { "epoch": 0.3472840605520926, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.347981452941895, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8584059476852417, "num_tokens": 104256553.0, "step": 2730 }, { "epoch": 0.3474112708306831, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.376805305480957, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.867268979549408, "num_tokens": 104291150.0, "step": 2731 }, { "epoch": 0.34753848110927366, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.393879890441895, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8462252616882324, "num_tokens": 104327977.0, "step": 2732 }, { "epoch": 0.34766569138786413, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.323958396911621, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8505853414535522, "num_tokens": 104365167.0, "step": 2733 }, { "epoch": 0.34779290166645466, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.380131721496582, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.859559178352356, "num_tokens": 104400304.0, "step": 2734 }, { "epoch": 0.3479201119450452, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.411540031433105, "learning_rate": 1e-06, "loss": 0.503, "mean_token_accuracy": 0.8382236957550049, "num_tokens": 104437974.0, "step": 2735 }, { "epoch": 0.34804732222363566, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.309089660644531, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8573460578918457, "num_tokens": 104474474.0, "step": 2736 }, { "epoch": 0.3481745325022262, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.427421569824219, "learning_rate": 1e-06, "loss": 0.5002, "mean_token_accuracy": 0.8392525911331177, "num_tokens": 104511004.0, "step": 2737 }, { "epoch": 0.3483017427808167, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.321478843688965, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8478049039840698, "num_tokens": 104551889.0, "step": 2738 }, { "epoch": 0.3484289530594072, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 9.331278800964355, "learning_rate": 1e-06, "loss": 0.5009, "mean_token_accuracy": 0.8378802537918091, "num_tokens": 104591962.0, "step": 2739 }, { "epoch": 0.3485561633379977, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.34846019744873, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8508847951889038, "num_tokens": 104636201.0, "step": 2740 }, { "epoch": 0.34868337361658824, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.2674560546875, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8533462285995483, "num_tokens": 104673448.0, "step": 2741 }, { "epoch": 0.3488105838951787, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 9.406716346740723, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8608360290527344, "num_tokens": 104708879.0, "step": 2742 }, { "epoch": 0.34893779417376924, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.310030937194824, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8641975522041321, "num_tokens": 104747073.0, "step": 2743 }, { "epoch": 0.3490650044523598, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.363648414611816, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8532814383506775, "num_tokens": 104786950.0, "step": 2744 }, { "epoch": 0.34919221473095025, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.379478454589844, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8590872287750244, "num_tokens": 104828306.0, "step": 2745 }, { "epoch": 0.3493194250095408, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.352434158325195, "learning_rate": 1e-06, "loss": 0.4913, "mean_token_accuracy": 0.8458061814308167, "num_tokens": 104867030.0, "step": 2746 }, { "epoch": 0.3494466352881313, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.339999198913574, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8584449291229248, "num_tokens": 104909182.0, "step": 2747 }, { "epoch": 0.3495738455667218, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.317740440368652, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.849533200263977, "num_tokens": 104950648.0, "step": 2748 }, { "epoch": 0.3497010558453123, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.317254066467285, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8591945767402649, "num_tokens": 104994780.0, "step": 2749 }, { "epoch": 0.34982826612390283, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.33362865447998, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8610751628875732, "num_tokens": 105029469.0, "step": 2750 }, { "epoch": 0.3499554764024933, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.324472427368164, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8479865193367004, "num_tokens": 105066797.0, "step": 2751 }, { "epoch": 0.35008268668108383, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.355920791625977, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8691602349281311, "num_tokens": 105100869.0, "step": 2752 }, { "epoch": 0.35020989695967436, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.364523887634277, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8663128614425659, "num_tokens": 105135409.0, "step": 2753 }, { "epoch": 0.35033710723826483, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.348920822143555, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8609687089920044, "num_tokens": 105172839.0, "step": 2754 }, { "epoch": 0.35046431751685536, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 9.333155632019043, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8599345684051514, "num_tokens": 105216656.0, "step": 2755 }, { "epoch": 0.3505915277954459, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.313769340515137, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8480969667434692, "num_tokens": 105253740.0, "step": 2756 }, { "epoch": 0.35071873807403636, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.433088302612305, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.858268141746521, "num_tokens": 105290073.0, "step": 2757 }, { "epoch": 0.3508459483526269, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.335107803344727, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.860636830329895, "num_tokens": 105325933.0, "step": 2758 }, { "epoch": 0.3509731586312174, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.41422176361084, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8577347993850708, "num_tokens": 105362962.0, "step": 2759 }, { "epoch": 0.3511003689098079, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.247833251953125, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8551056385040283, "num_tokens": 105410074.0, "step": 2760 }, { "epoch": 0.3512275791883984, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.374597549438477, "learning_rate": 1e-06, "loss": 0.5136, "mean_token_accuracy": 0.8395427465438843, "num_tokens": 105450264.0, "step": 2761 }, { "epoch": 0.35135478946698895, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8133392333984375e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.352237701416016, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8455063104629517, "num_tokens": 105486566.0, "step": 2762 }, { "epoch": 0.3514819997455794, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.273270606994629, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8580008745193481, "num_tokens": 105527185.0, "step": 2763 }, { "epoch": 0.35160921002416995, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.402247428894043, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8533418774604797, "num_tokens": 105557401.0, "step": 2764 }, { "epoch": 0.3517364203027605, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.335888862609863, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8609546422958374, "num_tokens": 105586877.0, "step": 2765 }, { "epoch": 0.35186363058135095, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.361760139465332, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8586925268173218, "num_tokens": 105626988.0, "step": 2766 }, { "epoch": 0.3519908408599415, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.278361320495605, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8454487323760986, "num_tokens": 105665796.0, "step": 2767 }, { "epoch": 0.352118051138532, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.33140754699707, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8456816673278809, "num_tokens": 105708892.0, "step": 2768 }, { "epoch": 0.3522452614171225, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.321375846862793, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8481618762016296, "num_tokens": 105748221.0, "step": 2769 }, { "epoch": 0.352372471695713, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.30860710144043, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8528486490249634, "num_tokens": 105784568.0, "step": 2770 }, { "epoch": 0.35249968197430354, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.34110164642334, "learning_rate": 1e-06, "loss": 0.5032, "mean_token_accuracy": 0.8423117995262146, "num_tokens": 105827223.0, "step": 2771 }, { "epoch": 0.352626892252894, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.33188533782959, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8724393844604492, "num_tokens": 105871014.0, "step": 2772 }, { "epoch": 0.35275410253148454, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.350056648254395, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8514354825019836, "num_tokens": 105908995.0, "step": 2773 }, { "epoch": 0.35288131281007507, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.331292152404785, "learning_rate": 1e-06, "loss": 0.4849, "mean_token_accuracy": 0.8482820391654968, "num_tokens": 105948243.0, "step": 2774 }, { "epoch": 0.35300852308866554, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.354458808898926, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8566538095474243, "num_tokens": 105983997.0, "step": 2775 }, { "epoch": 0.35313573336725607, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.86102294921875e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.395325660705566, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.8411712646484375, "num_tokens": 106020336.0, "step": 2776 }, { "epoch": 0.3532629436458466, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.335951805114746, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8561719059944153, "num_tokens": 106053300.0, "step": 2777 }, { "epoch": 0.35339015392443707, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.269462585449219, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8683664798736572, "num_tokens": 106091790.0, "step": 2778 }, { "epoch": 0.3535173642030276, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.428182601928711, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8573168516159058, "num_tokens": 106123768.0, "step": 2779 }, { "epoch": 0.3536445744816181, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.387761116027832, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8624363541603088, "num_tokens": 106158219.0, "step": 2780 }, { "epoch": 0.35377178476020865, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.371649742126465, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8493900299072266, "num_tokens": 106195795.0, "step": 2781 }, { "epoch": 0.3538989950387991, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.38478946685791, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8554580211639404, "num_tokens": 106234973.0, "step": 2782 }, { "epoch": 0.35402620531738965, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.425259590148926, "learning_rate": 1e-06, "loss": 0.5118, "mean_token_accuracy": 0.8407905101776123, "num_tokens": 106267297.0, "step": 2783 }, { "epoch": 0.3541534155959802, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.312068939208984, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8620444536209106, "num_tokens": 106309070.0, "step": 2784 }, { "epoch": 0.35428062587457065, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.324846267700195, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8580734729766846, "num_tokens": 106345784.0, "step": 2785 }, { "epoch": 0.3544078361531612, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.366573333740234, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8622509241104126, "num_tokens": 106383172.0, "step": 2786 }, { "epoch": 0.3545350464317517, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.367281913757324, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.852962076663971, "num_tokens": 106416916.0, "step": 2787 }, { "epoch": 0.3546622567103422, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.34383487701416, "learning_rate": 1e-06, "loss": 0.5044, "mean_token_accuracy": 0.8387836217880249, "num_tokens": 106456969.0, "step": 2788 }, { "epoch": 0.3547894669889327, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.402103424072266, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8521490097045898, "num_tokens": 106490476.0, "step": 2789 }, { "epoch": 0.35491667726752324, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.340797424316406, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8549270629882812, "num_tokens": 106525990.0, "step": 2790 }, { "epoch": 0.3550438875461137, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.362507820129395, "learning_rate": 1e-06, "loss": 0.5501, "mean_token_accuracy": 0.8294551968574524, "num_tokens": 106562953.0, "step": 2791 }, { "epoch": 0.35517109782470424, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.378925323486328, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8566468954086304, "num_tokens": 106599142.0, "step": 2792 }, { "epoch": 0.35529830810329477, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.379526138305664, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8664592504501343, "num_tokens": 106636743.0, "step": 2793 }, { "epoch": 0.35542551838188524, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.35579776763916, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8595799803733826, "num_tokens": 106678992.0, "step": 2794 }, { "epoch": 0.35555272866047577, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.346796035766602, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8452810049057007, "num_tokens": 106719884.0, "step": 2795 }, { "epoch": 0.3556799389390663, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.39771842956543, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8506657481193542, "num_tokens": 106755224.0, "step": 2796 }, { "epoch": 0.35580714921765677, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.383213996887207, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.866907000541687, "num_tokens": 106789366.0, "step": 2797 }, { "epoch": 0.3559343594962473, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.446955680847168, "learning_rate": 1e-06, "loss": 0.4807, "mean_token_accuracy": 0.8449777960777283, "num_tokens": 106826382.0, "step": 2798 }, { "epoch": 0.3560615697748378, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.332086563110352, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8452112674713135, "num_tokens": 106860027.0, "step": 2799 }, { "epoch": 0.3561887800534283, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.376197814941406, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8627710938453674, "num_tokens": 106898684.0, "step": 2800 }, { "epoch": 0.35631599033201883, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.432945251464844, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8563118577003479, "num_tokens": 106934496.0, "step": 2801 }, { "epoch": 0.35644320061060936, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.426431655883789, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.868140697479248, "num_tokens": 106974335.0, "step": 2802 }, { "epoch": 0.35657041088919983, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.366586685180664, "learning_rate": 1e-06, "loss": 0.5539, "mean_token_accuracy": 0.8295918703079224, "num_tokens": 107013914.0, "step": 2803 }, { "epoch": 0.35669762116779036, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.362176895141602, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.857732355594635, "num_tokens": 107053747.0, "step": 2804 }, { "epoch": 0.3568248314463809, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.366697311401367, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8438826203346252, "num_tokens": 107095245.0, "step": 2805 }, { "epoch": 0.35695204172497136, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.340895652770996, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8547058701515198, "num_tokens": 107133627.0, "step": 2806 }, { "epoch": 0.3570792520035619, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.370728492736816, "learning_rate": 1e-06, "loss": 0.5205, "mean_token_accuracy": 0.8294636011123657, "num_tokens": 107170128.0, "step": 2807 }, { "epoch": 0.3572064622821524, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.398104667663574, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8636851906776428, "num_tokens": 107209137.0, "step": 2808 }, { "epoch": 0.3573336725607429, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.396652221679688, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.849000871181488, "num_tokens": 107244866.0, "step": 2809 }, { "epoch": 0.3574608828393334, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.341313362121582, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8540458083152771, "num_tokens": 107287399.0, "step": 2810 }, { "epoch": 0.35758809311792394, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.35510540008545, "learning_rate": 1e-06, "loss": 0.4877, "mean_token_accuracy": 0.8441060781478882, "num_tokens": 107325627.0, "step": 2811 }, { "epoch": 0.3577153033965144, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.377716064453125, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8707736730575562, "num_tokens": 107359375.0, "step": 2812 }, { "epoch": 0.35784251367510495, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.427497863769531, "learning_rate": 1e-06, "loss": 0.4914, "mean_token_accuracy": 0.8426610231399536, "num_tokens": 107400286.0, "step": 2813 }, { "epoch": 0.3579697239536955, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.411137580871582, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8453035354614258, "num_tokens": 107431768.0, "step": 2814 }, { "epoch": 0.35809693423228595, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.369492530822754, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8633676767349243, "num_tokens": 107470875.0, "step": 2815 }, { "epoch": 0.3582241445108765, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.373434066772461, "learning_rate": 1e-06, "loss": 0.4815, "mean_token_accuracy": 0.8451781868934631, "num_tokens": 107510988.0, "step": 2816 }, { "epoch": 0.358351354789467, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.373427391052246, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8621009588241577, "num_tokens": 107552423.0, "step": 2817 }, { "epoch": 0.3584785650680575, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.395795822143555, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8469734191894531, "num_tokens": 107591693.0, "step": 2818 }, { "epoch": 0.358605775346648, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.374225616455078, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.859184741973877, "num_tokens": 107626017.0, "step": 2819 }, { "epoch": 0.35873298562523853, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.374959945678711, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8702754378318787, "num_tokens": 107664927.0, "step": 2820 }, { "epoch": 0.358860195903829, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.395840644836426, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8469870090484619, "num_tokens": 107702487.0, "step": 2821 }, { "epoch": 0.35898740618241953, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.377363204956055, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8671363592147827, "num_tokens": 107741263.0, "step": 2822 }, { "epoch": 0.35911461646101006, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.31696605682373, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8572170734405518, "num_tokens": 107781079.0, "step": 2823 }, { "epoch": 0.35924182673960053, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.386072158813477, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8726413249969482, "num_tokens": 107819309.0, "step": 2824 }, { "epoch": 0.35936903701819106, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.394698143005371, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8548935651779175, "num_tokens": 107853251.0, "step": 2825 }, { "epoch": 0.3594962472967816, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 9.33642292022705, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8741624355316162, "num_tokens": 107893476.0, "step": 2826 }, { "epoch": 0.35962345757537206, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.363953590393066, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8641616106033325, "num_tokens": 107931204.0, "step": 2827 }, { "epoch": 0.3597506678539626, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.438517570495605, "learning_rate": 1e-06, "loss": 0.5039, "mean_token_accuracy": 0.837230920791626, "num_tokens": 107966471.0, "step": 2828 }, { "epoch": 0.3598778781325531, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.372065544128418, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8554562330245972, "num_tokens": 108005328.0, "step": 2829 }, { "epoch": 0.36000508841114365, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.366852760314941, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8438132405281067, "num_tokens": 108046295.0, "step": 2830 }, { "epoch": 0.3601322986897341, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.43049430847168, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8562915921211243, "num_tokens": 108078164.0, "step": 2831 }, { "epoch": 0.36025950896832465, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.419629096984863, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8495283722877502, "num_tokens": 108117105.0, "step": 2832 }, { "epoch": 0.3603867192469152, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.376753807067871, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8521538972854614, "num_tokens": 108155301.0, "step": 2833 }, { "epoch": 0.36051392952550565, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.401562690734863, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8647913932800293, "num_tokens": 108193091.0, "step": 2834 }, { "epoch": 0.3606411398040962, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.379963874816895, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8527119755744934, "num_tokens": 108233719.0, "step": 2835 }, { "epoch": 0.3607683500826867, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.479413986206055, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8601394295692444, "num_tokens": 108265775.0, "step": 2836 }, { "epoch": 0.3608955603612772, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.312566757202148, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8645062446594238, "num_tokens": 108307953.0, "step": 2837 }, { "epoch": 0.3610227706398677, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.37421703338623, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8501059412956238, "num_tokens": 108351998.0, "step": 2838 }, { "epoch": 0.36114998091845824, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.40074634552002, "learning_rate": 1e-06, "loss": 0.4947, "mean_token_accuracy": 0.8430183529853821, "num_tokens": 108395968.0, "step": 2839 }, { "epoch": 0.3612771911970487, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.400053024291992, "learning_rate": 1e-06, "loss": 0.5174, "mean_token_accuracy": 0.8353472352027893, "num_tokens": 108433045.0, "step": 2840 }, { "epoch": 0.36140440147563924, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.403196334838867, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8489664793014526, "num_tokens": 108470964.0, "step": 2841 }, { "epoch": 0.36153161175422976, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.395333290100098, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8551901578903198, "num_tokens": 108512572.0, "step": 2842 }, { "epoch": 0.36165882203282024, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.401111602783203, "learning_rate": 1e-06, "loss": 0.4873, "mean_token_accuracy": 0.8405877351760864, "num_tokens": 108548526.0, "step": 2843 }, { "epoch": 0.36178603231141077, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.441205978393555, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8510688543319702, "num_tokens": 108588447.0, "step": 2844 }, { "epoch": 0.3619132425900013, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.369890213012695, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8447560667991638, "num_tokens": 108628426.0, "step": 2845 }, { "epoch": 0.36204045286859177, "ewc_loss": 0.01422119140625, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 9.410970687866211, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8492803573608398, "num_tokens": 108661355.0, "step": 2846 }, { "epoch": 0.3621676631471823, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.9325485229492188e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.43359088897705, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.870065450668335, "num_tokens": 108698505.0, "step": 2847 }, { "epoch": 0.3622948734257728, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.403977394104004, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8590054512023926, "num_tokens": 108737409.0, "step": 2848 }, { "epoch": 0.3624220837043633, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.41887378692627, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8553943634033203, "num_tokens": 108772969.0, "step": 2849 }, { "epoch": 0.3625492939829538, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 9.446982383728027, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8599249124526978, "num_tokens": 108805129.0, "step": 2850 }, { "epoch": 0.36267650426154435, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.379323959350586, "learning_rate": 1e-06, "loss": 0.4908, "mean_token_accuracy": 0.8406655788421631, "num_tokens": 108841197.0, "step": 2851 }, { "epoch": 0.3628037145401348, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.445657730102539, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8677718639373779, "num_tokens": 108876442.0, "step": 2852 }, { "epoch": 0.36293092481872535, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.427522659301758, "learning_rate": 1e-06, "loss": 0.4862, "mean_token_accuracy": 0.8401653170585632, "num_tokens": 108911432.0, "step": 2853 }, { "epoch": 0.3630581350973159, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.44714069366455, "learning_rate": 1e-06, "loss": 0.4981, "mean_token_accuracy": 0.8398311138153076, "num_tokens": 108952977.0, "step": 2854 }, { "epoch": 0.36318534537590635, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.357707023620605, "learning_rate": 1e-06, "loss": 0.5352, "mean_token_accuracy": 0.8310182690620422, "num_tokens": 108995602.0, "step": 2855 }, { "epoch": 0.3633125556544969, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.440816879272461, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8561438322067261, "num_tokens": 109040226.0, "step": 2856 }, { "epoch": 0.3634397659330874, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.389835357666016, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.856194257736206, "num_tokens": 109082898.0, "step": 2857 }, { "epoch": 0.3635669762116779, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.434281349182129, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8449660539627075, "num_tokens": 109122460.0, "step": 2858 }, { "epoch": 0.3636941864902684, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.403986930847168, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8535470962524414, "num_tokens": 109161680.0, "step": 2859 }, { "epoch": 0.36382139676885894, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.434854507446289, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8736951351165771, "num_tokens": 109198943.0, "step": 2860 }, { "epoch": 0.3639486070474494, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.44124984741211, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8620803356170654, "num_tokens": 109235719.0, "step": 2861 }, { "epoch": 0.36407581732603994, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.480144500732422, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8467106223106384, "num_tokens": 109272681.0, "step": 2862 }, { "epoch": 0.36420302760463047, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.395112991333008, "learning_rate": 1e-06, "loss": 0.4967, "mean_token_accuracy": 0.8434809446334839, "num_tokens": 109306114.0, "step": 2863 }, { "epoch": 0.36433023788322094, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.487068176269531, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8464576601982117, "num_tokens": 109337728.0, "step": 2864 }, { "epoch": 0.36445744816181147, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.40300464630127, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.844697892665863, "num_tokens": 109377809.0, "step": 2865 }, { "epoch": 0.364584658440402, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.456855773925781, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8683944940567017, "num_tokens": 109413499.0, "step": 2866 }, { "epoch": 0.36471186871899247, "ewc_loss": 0.0146484375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.427227973937988, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8444772958755493, "num_tokens": 109460961.0, "step": 2867 }, { "epoch": 0.364839078997583, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.401650428771973, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8647030591964722, "num_tokens": 109500615.0, "step": 2868 }, { "epoch": 0.36496628927617353, "ewc_loss": 0.0146484375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.44845199584961, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.856548547744751, "num_tokens": 109533921.0, "step": 2869 }, { "epoch": 0.365093499554764, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.417304992675781, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.844352126121521, "num_tokens": 109567069.0, "step": 2870 }, { "epoch": 0.36522070983335453, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.436219215393066, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8671279549598694, "num_tokens": 109603696.0, "step": 2871 }, { "epoch": 0.36534792011194506, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.406803131103516, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8720660209655762, "num_tokens": 109644590.0, "step": 2872 }, { "epoch": 0.36547513039053553, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.440018653869629, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8502021431922913, "num_tokens": 109678869.0, "step": 2873 }, { "epoch": 0.36560234066912606, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.440354347229004, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8488461971282959, "num_tokens": 109722517.0, "step": 2874 }, { "epoch": 0.3657295509477166, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.481965065002441, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8507514595985413, "num_tokens": 109757869.0, "step": 2875 }, { "epoch": 0.36585676122630706, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.487431526184082, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8491494059562683, "num_tokens": 109793162.0, "step": 2876 }, { "epoch": 0.3659839715048976, "ewc_loss": 0.0146484375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.4288330078125, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8608273267745972, "num_tokens": 109832942.0, "step": 2877 }, { "epoch": 0.3661111817834881, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.45898723602295, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8527313470840454, "num_tokens": 109876230.0, "step": 2878 }, { "epoch": 0.3662383920620786, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 9.418197631835938, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.844579815864563, "num_tokens": 109911844.0, "step": 2879 }, { "epoch": 0.3663656023406691, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.486385345458984, "learning_rate": 1e-06, "loss": 0.513, "mean_token_accuracy": 0.8426092863082886, "num_tokens": 109957619.0, "step": 2880 }, { "epoch": 0.36649281261925964, "ewc_loss": 0.0146484375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.40493392944336, "learning_rate": 1e-06, "loss": 0.4832, "mean_token_accuracy": 0.8435044288635254, "num_tokens": 109999117.0, "step": 2881 }, { "epoch": 0.3666200228978502, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.4266996383667, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8611268997192383, "num_tokens": 110039557.0, "step": 2882 }, { "epoch": 0.36674723317644065, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.46187973022461, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8456606864929199, "num_tokens": 110074418.0, "step": 2883 }, { "epoch": 0.3668744434550312, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.442538261413574, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8590341806411743, "num_tokens": 110109255.0, "step": 2884 }, { "epoch": 0.3670016537336217, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.426698684692383, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8604357242584229, "num_tokens": 110150180.0, "step": 2885 }, { "epoch": 0.3671288640122122, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.418622016906738, "learning_rate": 1e-06, "loss": 0.507, "mean_token_accuracy": 0.837696373462677, "num_tokens": 110193653.0, "step": 2886 }, { "epoch": 0.3672560742908027, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.498526573181152, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8613154888153076, "num_tokens": 110231020.0, "step": 2887 }, { "epoch": 0.36738328456939323, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.38884449005127, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8696302175521851, "num_tokens": 110269397.0, "step": 2888 }, { "epoch": 0.3675104948479837, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.418400764465332, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8688350915908813, "num_tokens": 110311188.0, "step": 2889 }, { "epoch": 0.36763770512657423, "ewc_loss": 0.0146484375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.477940559387207, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8586834669113159, "num_tokens": 110343801.0, "step": 2890 }, { "epoch": 0.36776491540516476, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.438761711120605, "learning_rate": 1e-06, "loss": 0.503, "mean_token_accuracy": 0.8382601141929626, "num_tokens": 110385830.0, "step": 2891 }, { "epoch": 0.36789212568375523, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.507739067077637, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8542537093162537, "num_tokens": 110421494.0, "step": 2892 }, { "epoch": 0.36801933596234576, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.394341468811035, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8533320426940918, "num_tokens": 110463132.0, "step": 2893 }, { "epoch": 0.3681465462409363, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 9.466179847717285, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.856763482093811, "num_tokens": 110498330.0, "step": 2894 }, { "epoch": 0.36827375651952676, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.428165435791016, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8624224662780762, "num_tokens": 110534438.0, "step": 2895 }, { "epoch": 0.3684009667981173, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.440839767456055, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8496907949447632, "num_tokens": 110572059.0, "step": 2896 }, { "epoch": 0.3685281770767078, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.47863483428955, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.856359601020813, "num_tokens": 110611012.0, "step": 2897 }, { "epoch": 0.3686553873552983, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.439178466796875, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8615463376045227, "num_tokens": 110647429.0, "step": 2898 }, { "epoch": 0.3687825976338888, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.425267219543457, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8492765426635742, "num_tokens": 110685052.0, "step": 2899 }, { "epoch": 0.36890980791247935, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.420549392700195, "learning_rate": 1e-06, "loss": 0.5078, "mean_token_accuracy": 0.8335434198379517, "num_tokens": 110725222.0, "step": 2900 }, { "epoch": 0.3690370181910698, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.415718078613281, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8433890342712402, "num_tokens": 110764314.0, "step": 2901 }, { "epoch": 0.36916422846966035, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.393793106079102, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8655019402503967, "num_tokens": 110803585.0, "step": 2902 }, { "epoch": 0.3692914387482509, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.404975891113281, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8490970134735107, "num_tokens": 110846532.0, "step": 2903 }, { "epoch": 0.36941864902684135, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.421952247619629, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.85175621509552, "num_tokens": 110889807.0, "step": 2904 }, { "epoch": 0.3695458593054319, "ewc_loss": 0.01458740234375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.4013671875, "learning_rate": 1e-06, "loss": 0.5109, "mean_token_accuracy": 0.8353040218353271, "num_tokens": 110930567.0, "step": 2905 }, { "epoch": 0.3696730695840224, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.492005348205566, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8610879778862, "num_tokens": 110968725.0, "step": 2906 }, { "epoch": 0.3698002798626129, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.414796829223633, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.877843976020813, "num_tokens": 111002724.0, "step": 2907 }, { "epoch": 0.3699274901412034, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.49203872680664, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.85311359167099, "num_tokens": 111036668.0, "step": 2908 }, { "epoch": 0.37005470041979394, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.459800720214844, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8447951078414917, "num_tokens": 111077712.0, "step": 2909 }, { "epoch": 0.3701819106983844, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.390786170959473, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8449772596359253, "num_tokens": 111118856.0, "step": 2910 }, { "epoch": 0.37030912097697494, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.431138038635254, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8453850746154785, "num_tokens": 111156712.0, "step": 2911 }, { "epoch": 0.37043633125556547, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.460427284240723, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8553104400634766, "num_tokens": 111195542.0, "step": 2912 }, { "epoch": 0.37056354153415594, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.458014488220215, "learning_rate": 1e-06, "loss": 0.5255, "mean_token_accuracy": 0.8362486362457275, "num_tokens": 111237827.0, "step": 2913 }, { "epoch": 0.37069075181274647, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.431811332702637, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8622523546218872, "num_tokens": 111278802.0, "step": 2914 }, { "epoch": 0.370817962091337, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.468955993652344, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8517086505889893, "num_tokens": 111313795.0, "step": 2915 }, { "epoch": 0.37094517236992747, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.404306411743164, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8607336282730103, "num_tokens": 111353798.0, "step": 2916 }, { "epoch": 0.371072382648518, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.463480949401855, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8539736270904541, "num_tokens": 111390613.0, "step": 2917 }, { "epoch": 0.3711995929271085, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.455196380615234, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8543496131896973, "num_tokens": 111432043.0, "step": 2918 }, { "epoch": 0.371326803205699, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.388860702514648, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8606026768684387, "num_tokens": 111476575.0, "step": 2919 }, { "epoch": 0.3714540134842895, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.470124244689941, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8543577194213867, "num_tokens": 111515304.0, "step": 2920 }, { "epoch": 0.37158122376288005, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.471091270446777, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8640020489692688, "num_tokens": 111550246.0, "step": 2921 }, { "epoch": 0.3717084340414705, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.447617530822754, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8552244901657104, "num_tokens": 111592189.0, "step": 2922 }, { "epoch": 0.37183564432006105, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.461064338684082, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8574967384338379, "num_tokens": 111638177.0, "step": 2923 }, { "epoch": 0.3719628545986516, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.460692405700684, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.856027364730835, "num_tokens": 111674365.0, "step": 2924 }, { "epoch": 0.37209006487724205, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.485671997070312, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8493658304214478, "num_tokens": 111714571.0, "step": 2925 }, { "epoch": 0.3722172751558326, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.439974784851074, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8546793460845947, "num_tokens": 111755214.0, "step": 2926 }, { "epoch": 0.3723444854344231, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 9.437322616577148, "learning_rate": 1e-06, "loss": 0.5364, "mean_token_accuracy": 0.8275494575500488, "num_tokens": 111796516.0, "step": 2927 }, { "epoch": 0.3724716957130136, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.464077949523926, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.8399295210838318, "num_tokens": 111832615.0, "step": 2928 }, { "epoch": 0.3725989059916041, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.403368949890137, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.844893217086792, "num_tokens": 111874682.0, "step": 2929 }, { "epoch": 0.37272611627019464, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 9.45142650604248, "learning_rate": 1e-06, "loss": 0.5027, "mean_token_accuracy": 0.8422194719314575, "num_tokens": 111916369.0, "step": 2930 }, { "epoch": 0.37285332654878517, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.452896118164062, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8458923101425171, "num_tokens": 111950671.0, "step": 2931 }, { "epoch": 0.37298053682737564, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.440852165222168, "learning_rate": 1e-06, "loss": 0.5027, "mean_token_accuracy": 0.8357460498809814, "num_tokens": 111988846.0, "step": 2932 }, { "epoch": 0.37310774710596617, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.441883087158203, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8635748624801636, "num_tokens": 112022853.0, "step": 2933 }, { "epoch": 0.3732349573845567, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.443382263183594, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8501135110855103, "num_tokens": 112065631.0, "step": 2934 }, { "epoch": 0.37336216766314717, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.418320655822754, "learning_rate": 1e-06, "loss": 0.4893, "mean_token_accuracy": 0.8437345623970032, "num_tokens": 112107236.0, "step": 2935 }, { "epoch": 0.3734893779417377, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.442116737365723, "learning_rate": 1e-06, "loss": 0.5174, "mean_token_accuracy": 0.8315294981002808, "num_tokens": 112153914.0, "step": 2936 }, { "epoch": 0.3736165882203282, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.495499610900879, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.855866014957428, "num_tokens": 112187596.0, "step": 2937 }, { "epoch": 0.3737437984989187, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.534212112426758, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8520845770835876, "num_tokens": 112222409.0, "step": 2938 }, { "epoch": 0.37387100877750923, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.457615852355957, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8411460518836975, "num_tokens": 112257699.0, "step": 2939 }, { "epoch": 0.37399821905609976, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.562612533569336, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8623846173286438, "num_tokens": 112292431.0, "step": 2940 }, { "epoch": 0.37412542933469023, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.461076736450195, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8544137477874756, "num_tokens": 112331241.0, "step": 2941 }, { "epoch": 0.37425263961328076, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.474771499633789, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.849934458732605, "num_tokens": 112368486.0, "step": 2942 }, { "epoch": 0.3743798498918713, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.521648406982422, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8440335988998413, "num_tokens": 112409638.0, "step": 2943 }, { "epoch": 0.37450706017046176, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.415338516235352, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8485207557678223, "num_tokens": 112448257.0, "step": 2944 }, { "epoch": 0.3746342704490523, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.455842971801758, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8682106733322144, "num_tokens": 112488739.0, "step": 2945 }, { "epoch": 0.3747614807276428, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.462114334106445, "learning_rate": 1e-06, "loss": 0.4996, "mean_token_accuracy": 0.8388285636901855, "num_tokens": 112528621.0, "step": 2946 }, { "epoch": 0.3748886910062333, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.48926067352295, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8480001091957092, "num_tokens": 112564615.0, "step": 2947 }, { "epoch": 0.3750159012848238, "ewc_loss": 0.0146484375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 9.472366333007812, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8424069285392761, "num_tokens": 112598190.0, "step": 2948 }, { "epoch": 0.37514311156341434, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.444355964660645, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8524919748306274, "num_tokens": 112632078.0, "step": 2949 }, { "epoch": 0.3752703218420048, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.501466751098633, "learning_rate": 1e-06, "loss": 0.5268, "mean_token_accuracy": 0.830466628074646, "num_tokens": 112667848.0, "step": 2950 }, { "epoch": 0.37539753212059535, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.428638458251953, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8519114255905151, "num_tokens": 112707037.0, "step": 2951 }, { "epoch": 0.3755247423991859, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.464559555053711, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8619104027748108, "num_tokens": 112751921.0, "step": 2952 }, { "epoch": 0.37565195267777635, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.482230186462402, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.857596218585968, "num_tokens": 112793061.0, "step": 2953 }, { "epoch": 0.3757791629563669, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.445120811462402, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8628589510917664, "num_tokens": 112833852.0, "step": 2954 }, { "epoch": 0.3759063732349574, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.504674911499023, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8498085737228394, "num_tokens": 112866718.0, "step": 2955 }, { "epoch": 0.3760335835135479, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.455860137939453, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8542864322662354, "num_tokens": 112909420.0, "step": 2956 }, { "epoch": 0.3761607937921384, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.504216194152832, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8565757274627686, "num_tokens": 112944570.0, "step": 2957 }, { "epoch": 0.37628800407072893, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.449880599975586, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8655761480331421, "num_tokens": 112984107.0, "step": 2958 }, { "epoch": 0.3764152143493194, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.467728614807129, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8764153718948364, "num_tokens": 113023066.0, "step": 2959 }, { "epoch": 0.37654242462790993, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.498869895935059, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8517817854881287, "num_tokens": 113060093.0, "step": 2960 }, { "epoch": 0.37666963490650046, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 9.446882247924805, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8452714085578918, "num_tokens": 113095940.0, "step": 2961 }, { "epoch": 0.37679684518509093, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.506407737731934, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8643233776092529, "num_tokens": 113130699.0, "step": 2962 }, { "epoch": 0.37692405546368146, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.445662498474121, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8532705307006836, "num_tokens": 113169145.0, "step": 2963 }, { "epoch": 0.377051265742272, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.502952575683594, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8686009049415588, "num_tokens": 113201987.0, "step": 2964 }, { "epoch": 0.37717847602086246, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.444578170776367, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8598084449768066, "num_tokens": 113239588.0, "step": 2965 }, { "epoch": 0.377305686299453, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.459131240844727, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8555541038513184, "num_tokens": 113277075.0, "step": 2966 }, { "epoch": 0.3774328965780435, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.465028762817383, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.881187915802002, "num_tokens": 113313708.0, "step": 2967 }, { "epoch": 0.377560106856634, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.502830505371094, "learning_rate": 1e-06, "loss": 0.5093, "mean_token_accuracy": 0.8362102508544922, "num_tokens": 113344027.0, "step": 2968 }, { "epoch": 0.3776873171352245, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.456400871276855, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.858403742313385, "num_tokens": 113388444.0, "step": 2969 }, { "epoch": 0.37781452741381505, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.4314603805542, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8494364023208618, "num_tokens": 113428660.0, "step": 2970 }, { "epoch": 0.3779417376924055, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.482986450195312, "learning_rate": 1e-06, "loss": 0.5195, "mean_token_accuracy": 0.8322435617446899, "num_tokens": 113469160.0, "step": 2971 }, { "epoch": 0.37806894797099605, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.455766677856445, "learning_rate": 1e-06, "loss": 0.5208, "mean_token_accuracy": 0.8297341465950012, "num_tokens": 113504291.0, "step": 2972 }, { "epoch": 0.3781961582495866, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.449624061584473, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8560585975646973, "num_tokens": 113548085.0, "step": 2973 }, { "epoch": 0.37832336852817705, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.462010383605957, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8584740161895752, "num_tokens": 113581664.0, "step": 2974 }, { "epoch": 0.3784505788067676, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.461052894592285, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8670580387115479, "num_tokens": 113618552.0, "step": 2975 }, { "epoch": 0.3785777890853581, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.46654987335205, "learning_rate": 1e-06, "loss": 0.5267, "mean_token_accuracy": 0.8337814807891846, "num_tokens": 113653406.0, "step": 2976 }, { "epoch": 0.3787049993639486, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.457446098327637, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.862238883972168, "num_tokens": 113693803.0, "step": 2977 }, { "epoch": 0.3788322096425391, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.422140121459961, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8679708242416382, "num_tokens": 113734637.0, "step": 2978 }, { "epoch": 0.37895941992112964, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.4589204788208, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.846808671951294, "num_tokens": 113776579.0, "step": 2979 }, { "epoch": 0.37908663019972016, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.411213874816895, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8430182933807373, "num_tokens": 113818857.0, "step": 2980 }, { "epoch": 0.37921384047831064, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.503600120544434, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8454281091690063, "num_tokens": 113853248.0, "step": 2981 }, { "epoch": 0.37934105075690117, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.458476066589355, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8599472641944885, "num_tokens": 113894206.0, "step": 2982 }, { "epoch": 0.3794682610354917, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.46981143951416, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8611924648284912, "num_tokens": 113931309.0, "step": 2983 }, { "epoch": 0.37959547131408217, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.463486671447754, "learning_rate": 1e-06, "loss": 0.4815, "mean_token_accuracy": 0.8433742523193359, "num_tokens": 113976559.0, "step": 2984 }, { "epoch": 0.3797226815926727, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.489385604858398, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8525345921516418, "num_tokens": 114010814.0, "step": 2985 }, { "epoch": 0.3798498918712632, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.439291000366211, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8721717596054077, "num_tokens": 114049389.0, "step": 2986 }, { "epoch": 0.3799771021498537, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.477575302124023, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8716088533401489, "num_tokens": 114081463.0, "step": 2987 }, { "epoch": 0.3801043124284442, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.439335823059082, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8696417808532715, "num_tokens": 114117175.0, "step": 2988 }, { "epoch": 0.38023152270703475, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.474119186401367, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.8427362442016602, "num_tokens": 114153764.0, "step": 2989 }, { "epoch": 0.3803587329856252, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.501914024353027, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8534528017044067, "num_tokens": 114191487.0, "step": 2990 }, { "epoch": 0.38048594326421575, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.510852813720703, "learning_rate": 1e-06, "loss": 0.5235, "mean_token_accuracy": 0.8289851546287537, "num_tokens": 114233880.0, "step": 2991 }, { "epoch": 0.3806131535428063, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.507451057434082, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8517087697982788, "num_tokens": 114269238.0, "step": 2992 }, { "epoch": 0.38074036382139675, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.470081329345703, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8654945492744446, "num_tokens": 114307645.0, "step": 2993 }, { "epoch": 0.3808675740999873, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.558804512023926, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8573877215385437, "num_tokens": 114341800.0, "step": 2994 }, { "epoch": 0.3809947843785778, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.450657844543457, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8620916604995728, "num_tokens": 114378363.0, "step": 2995 }, { "epoch": 0.3811219946571683, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.50757122039795, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.8426607847213745, "num_tokens": 114414489.0, "step": 2996 }, { "epoch": 0.3812492049357588, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.50028133392334, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8489680290222168, "num_tokens": 114449197.0, "step": 2997 }, { "epoch": 0.38137641521434934, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.510852813720703, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8505107164382935, "num_tokens": 114478654.0, "step": 2998 }, { "epoch": 0.3815036254929398, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.463629722595215, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8626911640167236, "num_tokens": 114521807.0, "step": 2999 }, { "epoch": 0.38163083577153034, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.461196899414062, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8671196699142456, "num_tokens": 114558297.0, "step": 3000 }, { "epoch": 0.38175804605012087, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.487199783325195, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.860255777835846, "num_tokens": 114599630.0, "step": 3001 }, { "epoch": 0.38188525632871134, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.481389999389648, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8516523241996765, "num_tokens": 114638697.0, "step": 3002 }, { "epoch": 0.38201246660730187, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.46321964263916, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.85495924949646, "num_tokens": 114676106.0, "step": 3003 }, { "epoch": 0.3821396768858924, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.470226287841797, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8526421785354614, "num_tokens": 114712026.0, "step": 3004 }, { "epoch": 0.38226688716448287, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.465720176696777, "learning_rate": 1e-06, "loss": 0.5173, "mean_token_accuracy": 0.8328162431716919, "num_tokens": 114752348.0, "step": 3005 }, { "epoch": 0.3823940974430734, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.449055671691895, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8581400513648987, "num_tokens": 114789922.0, "step": 3006 }, { "epoch": 0.38252130772166393, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.46967887878418, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.848050057888031, "num_tokens": 114838834.0, "step": 3007 }, { "epoch": 0.3826485180002544, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.46595573425293, "learning_rate": 1e-06, "loss": 0.5066, "mean_token_accuracy": 0.8339421153068542, "num_tokens": 114873320.0, "step": 3008 }, { "epoch": 0.38277572827884493, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.468584060668945, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8662592172622681, "num_tokens": 114908293.0, "step": 3009 }, { "epoch": 0.38290293855743546, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.488210678100586, "learning_rate": 1e-06, "loss": 0.4825, "mean_token_accuracy": 0.8447734713554382, "num_tokens": 114950083.0, "step": 3010 }, { "epoch": 0.38303014883602593, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.52926254272461, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8589521646499634, "num_tokens": 114982353.0, "step": 3011 }, { "epoch": 0.38315735911461646, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.509147644042969, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8602441549301147, "num_tokens": 115017668.0, "step": 3012 }, { "epoch": 0.383284569393207, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.482231140136719, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.865869402885437, "num_tokens": 115055266.0, "step": 3013 }, { "epoch": 0.38341177967179746, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.493791580200195, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8494317531585693, "num_tokens": 115089728.0, "step": 3014 }, { "epoch": 0.383538989950388, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.479754447937012, "learning_rate": 1e-06, "loss": 0.4929, "mean_token_accuracy": 0.8387027978897095, "num_tokens": 115135192.0, "step": 3015 }, { "epoch": 0.3836662002289785, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.555378913879395, "learning_rate": 1e-06, "loss": 0.5096, "mean_token_accuracy": 0.8363195061683655, "num_tokens": 115170350.0, "step": 3016 }, { "epoch": 0.383793410507569, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.492305755615234, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8603869676589966, "num_tokens": 115202724.0, "step": 3017 }, { "epoch": 0.3839206207861595, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.50214672088623, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8537516593933105, "num_tokens": 115239274.0, "step": 3018 }, { "epoch": 0.38404783106475004, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.51502799987793, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8624637126922607, "num_tokens": 115278330.0, "step": 3019 }, { "epoch": 0.3841750413433405, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.478219032287598, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.848949670791626, "num_tokens": 115321258.0, "step": 3020 }, { "epoch": 0.38430225162193105, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.492719650268555, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8555623888969421, "num_tokens": 115362609.0, "step": 3021 }, { "epoch": 0.3844294619005216, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.496407508850098, "learning_rate": 1e-06, "loss": 0.543, "mean_token_accuracy": 0.827779233455658, "num_tokens": 115405814.0, "step": 3022 }, { "epoch": 0.38455667217911205, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.531229972839355, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8472869396209717, "num_tokens": 115440856.0, "step": 3023 }, { "epoch": 0.3846838824577026, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.472527503967285, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8600116968154907, "num_tokens": 115482582.0, "step": 3024 }, { "epoch": 0.3848110927362931, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.500970840454102, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8451569080352783, "num_tokens": 115527268.0, "step": 3025 }, { "epoch": 0.3849383030148836, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.477089881896973, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8486223220825195, "num_tokens": 115563757.0, "step": 3026 }, { "epoch": 0.3850655132934741, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.4628267288208, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8579336404800415, "num_tokens": 115599629.0, "step": 3027 }, { "epoch": 0.38519272357206463, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.504947662353516, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8582245111465454, "num_tokens": 115634261.0, "step": 3028 }, { "epoch": 0.3853199338506551, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.454989433288574, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8667610287666321, "num_tokens": 115671150.0, "step": 3029 }, { "epoch": 0.38544714412924563, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.492302894592285, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8669314384460449, "num_tokens": 115714328.0, "step": 3030 }, { "epoch": 0.38557435440783616, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.502057075500488, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8695776462554932, "num_tokens": 115747118.0, "step": 3031 }, { "epoch": 0.3857015646864267, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.505693435668945, "learning_rate": 1e-06, "loss": 0.516, "mean_token_accuracy": 0.8340357542037964, "num_tokens": 115790309.0, "step": 3032 }, { "epoch": 0.38582877496501716, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.490203857421875, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8664186000823975, "num_tokens": 115825630.0, "step": 3033 }, { "epoch": 0.3859559852436077, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.557512283325195, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8616591691970825, "num_tokens": 115858848.0, "step": 3034 }, { "epoch": 0.3860831955221982, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.488536834716797, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8632110953330994, "num_tokens": 115900024.0, "step": 3035 }, { "epoch": 0.3862104058007887, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.564315795898438, "learning_rate": 1e-06, "loss": 0.5285, "mean_token_accuracy": 0.8293176293373108, "num_tokens": 115938899.0, "step": 3036 }, { "epoch": 0.3863376160793792, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.540653228759766, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8619186282157898, "num_tokens": 115972199.0, "step": 3037 }, { "epoch": 0.38646482635796975, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.475231170654297, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8644235134124756, "num_tokens": 116011506.0, "step": 3038 }, { "epoch": 0.3865920366365602, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.513534545898438, "learning_rate": 1e-06, "loss": 0.5274, "mean_token_accuracy": 0.8312865495681763, "num_tokens": 116051428.0, "step": 3039 }, { "epoch": 0.38671924691515075, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.497469902038574, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.863103985786438, "num_tokens": 116090307.0, "step": 3040 }, { "epoch": 0.3868464571937413, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.515591621398926, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.844623327255249, "num_tokens": 116123381.0, "step": 3041 }, { "epoch": 0.38697366747233175, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.571090698242188, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8568528890609741, "num_tokens": 116157755.0, "step": 3042 }, { "epoch": 0.3871008777509223, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.48892593383789, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8549093008041382, "num_tokens": 116197371.0, "step": 3043 }, { "epoch": 0.3872280880295128, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.531137466430664, "learning_rate": 1e-06, "loss": 0.5042, "mean_token_accuracy": 0.8398025035858154, "num_tokens": 116240544.0, "step": 3044 }, { "epoch": 0.3873552983081033, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.550870895385742, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8519084453582764, "num_tokens": 116272079.0, "step": 3045 }, { "epoch": 0.3874825085866938, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.533648490905762, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8517090678215027, "num_tokens": 116309641.0, "step": 3046 }, { "epoch": 0.38760971886528434, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.521700859069824, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8494495749473572, "num_tokens": 116344470.0, "step": 3047 }, { "epoch": 0.3877369291438748, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.541563987731934, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8581389784812927, "num_tokens": 116380177.0, "step": 3048 }, { "epoch": 0.38786413942246534, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.521774291992188, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8645771145820618, "num_tokens": 116416482.0, "step": 3049 }, { "epoch": 0.38799134970105587, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.472650527954102, "learning_rate": 1e-06, "loss": 0.4977, "mean_token_accuracy": 0.8414944410324097, "num_tokens": 116456103.0, "step": 3050 }, { "epoch": 0.38811855997964634, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.547093391418457, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8597425818443298, "num_tokens": 116496810.0, "step": 3051 }, { "epoch": 0.38824577025823687, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.552417755126953, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8508920669555664, "num_tokens": 116532999.0, "step": 3052 }, { "epoch": 0.3883729805368274, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.514459609985352, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8628662824630737, "num_tokens": 116565700.0, "step": 3053 }, { "epoch": 0.38850019081541787, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.471625328063965, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8583434224128723, "num_tokens": 116602798.0, "step": 3054 }, { "epoch": 0.3886274010940084, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.523919105529785, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8644524812698364, "num_tokens": 116638743.0, "step": 3055 }, { "epoch": 0.3887546113725989, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.503462791442871, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8671712875366211, "num_tokens": 116672290.0, "step": 3056 }, { "epoch": 0.3888818216511894, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.513813018798828, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8560667634010315, "num_tokens": 116712593.0, "step": 3057 }, { "epoch": 0.3890090319297799, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.56277847290039, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8529889583587646, "num_tokens": 116748663.0, "step": 3058 }, { "epoch": 0.38913624220837045, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.532793045043945, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8514305353164673, "num_tokens": 116796018.0, "step": 3059 }, { "epoch": 0.3892634524869609, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.511863708496094, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8517778515815735, "num_tokens": 116837287.0, "step": 3060 }, { "epoch": 0.38939066276555145, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.587356567382812, "learning_rate": 1e-06, "loss": 0.5206, "mean_token_accuracy": 0.8383181095123291, "num_tokens": 116872163.0, "step": 3061 }, { "epoch": 0.389517873044142, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.57797622680664, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.863106906414032, "num_tokens": 116913445.0, "step": 3062 }, { "epoch": 0.38964508332273246, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.48814868927002, "learning_rate": 1e-06, "loss": 0.4823, "mean_token_accuracy": 0.844759464263916, "num_tokens": 116954469.0, "step": 3063 }, { "epoch": 0.389772293601323, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.556483268737793, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.854869544506073, "num_tokens": 116994339.0, "step": 3064 }, { "epoch": 0.3898995038799135, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.582231521606445, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8666888475418091, "num_tokens": 117030867.0, "step": 3065 }, { "epoch": 0.390026714158504, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.519174575805664, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8627083897590637, "num_tokens": 117068309.0, "step": 3066 }, { "epoch": 0.3901539244370945, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.57329273223877, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8503459692001343, "num_tokens": 117105278.0, "step": 3067 }, { "epoch": 0.39028113471568504, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.54916763305664, "learning_rate": 1e-06, "loss": 0.4893, "mean_token_accuracy": 0.8462954759597778, "num_tokens": 117143096.0, "step": 3068 }, { "epoch": 0.3904083449942755, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.526809692382812, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8552957773208618, "num_tokens": 117179638.0, "step": 3069 }, { "epoch": 0.39053555527286604, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.604035377502441, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8670091032981873, "num_tokens": 117211560.0, "step": 3070 }, { "epoch": 0.39066276555145657, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.552794456481934, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8462274074554443, "num_tokens": 117250622.0, "step": 3071 }, { "epoch": 0.39078997583004704, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.563676834106445, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8745859861373901, "num_tokens": 117287600.0, "step": 3072 }, { "epoch": 0.39091718610863757, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.560454368591309, "learning_rate": 1e-06, "loss": 0.4994, "mean_token_accuracy": 0.8393291234970093, "num_tokens": 117328188.0, "step": 3073 }, { "epoch": 0.3910443963872281, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.574066162109375, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8630848526954651, "num_tokens": 117367282.0, "step": 3074 }, { "epoch": 0.39117160666581857, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.558295249938965, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8661472797393799, "num_tokens": 117403376.0, "step": 3075 }, { "epoch": 0.3912988169444091, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.557957649230957, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8455222845077515, "num_tokens": 117438661.0, "step": 3076 }, { "epoch": 0.39142602722299963, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.612245559692383, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8597186803817749, "num_tokens": 117477351.0, "step": 3077 }, { "epoch": 0.3915532375015901, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.535825729370117, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8514115214347839, "num_tokens": 117512773.0, "step": 3078 }, { "epoch": 0.39168044778018063, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.599733352661133, "learning_rate": 1e-06, "loss": 0.4996, "mean_token_accuracy": 0.8398860692977905, "num_tokens": 117555370.0, "step": 3079 }, { "epoch": 0.39180765805877116, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.518604278564453, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8623257875442505, "num_tokens": 117594208.0, "step": 3080 }, { "epoch": 0.3919348683373617, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.515949249267578, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8578446507453918, "num_tokens": 117632715.0, "step": 3081 }, { "epoch": 0.39206207861595216, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.554895401000977, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8499587178230286, "num_tokens": 117669221.0, "step": 3082 }, { "epoch": 0.3921892888945427, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.517267227172852, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8487067818641663, "num_tokens": 117709559.0, "step": 3083 }, { "epoch": 0.3923164991731332, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.53132152557373, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8619402647018433, "num_tokens": 117745240.0, "step": 3084 }, { "epoch": 0.3924437094517237, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.49891471862793, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8568583726882935, "num_tokens": 117787897.0, "step": 3085 }, { "epoch": 0.3925709197303142, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.54136848449707, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8506044149398804, "num_tokens": 117826862.0, "step": 3086 }, { "epoch": 0.39269813000890474, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.522972106933594, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8495086431503296, "num_tokens": 117866932.0, "step": 3087 }, { "epoch": 0.3928253402874952, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.524408340454102, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8554083704948425, "num_tokens": 117905538.0, "step": 3088 }, { "epoch": 0.39295255056608575, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.531365394592285, "learning_rate": 1e-06, "loss": 0.4933, "mean_token_accuracy": 0.8428387641906738, "num_tokens": 117945123.0, "step": 3089 }, { "epoch": 0.3930797608446763, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.519174575805664, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8574590682983398, "num_tokens": 117988743.0, "step": 3090 }, { "epoch": 0.39320697112326675, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.473555564880371, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8694381713867188, "num_tokens": 118026078.0, "step": 3091 }, { "epoch": 0.3933341814018573, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.604623794555664, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8660897016525269, "num_tokens": 118060105.0, "step": 3092 }, { "epoch": 0.3934613916804478, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.530380249023438, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8825769424438477, "num_tokens": 118089677.0, "step": 3093 }, { "epoch": 0.3935886019590383, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.51741886138916, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8682763576507568, "num_tokens": 118128449.0, "step": 3094 }, { "epoch": 0.3937158122376288, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.490527153015137, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8636636734008789, "num_tokens": 118168411.0, "step": 3095 }, { "epoch": 0.39384302251621933, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.553159713745117, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.865593433380127, "num_tokens": 118206097.0, "step": 3096 }, { "epoch": 0.3939702327948098, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.47146224975586, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8683344125747681, "num_tokens": 118244429.0, "step": 3097 }, { "epoch": 0.39409744307340033, "ewc_loss": 0.014892578125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 9.600997924804688, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8613154292106628, "num_tokens": 118279125.0, "step": 3098 }, { "epoch": 0.39422465335199086, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.518989562988281, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8528410196304321, "num_tokens": 118313156.0, "step": 3099 }, { "epoch": 0.39435186363058133, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.560389518737793, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8419773578643799, "num_tokens": 118349868.0, "step": 3100 }, { "epoch": 0.39447907390917186, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.533019065856934, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8456951379776001, "num_tokens": 118394998.0, "step": 3101 }, { "epoch": 0.3946062841877624, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.570389747619629, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8670430779457092, "num_tokens": 118436002.0, "step": 3102 }, { "epoch": 0.39473349446635286, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.524394035339355, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8602330684661865, "num_tokens": 118475094.0, "step": 3103 }, { "epoch": 0.3948607047449434, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.539488792419434, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8523364067077637, "num_tokens": 118514164.0, "step": 3104 }, { "epoch": 0.3949879150235339, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 9.535333633422852, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8514934778213501, "num_tokens": 118554271.0, "step": 3105 }, { "epoch": 0.3951151253021244, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.50823974609375, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.856264591217041, "num_tokens": 118597123.0, "step": 3106 }, { "epoch": 0.3952423355807149, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.571483612060547, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.849638819694519, "num_tokens": 118636683.0, "step": 3107 }, { "epoch": 0.39536954585930545, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.582289695739746, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8590378761291504, "num_tokens": 118665770.0, "step": 3108 }, { "epoch": 0.3954967561378959, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.502503395080566, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8664206266403198, "num_tokens": 118701652.0, "step": 3109 }, { "epoch": 0.39562396641648645, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.564821243286133, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8613448143005371, "num_tokens": 118741756.0, "step": 3110 }, { "epoch": 0.395751176695077, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.536713600158691, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8566539883613586, "num_tokens": 118784320.0, "step": 3111 }, { "epoch": 0.39587838697366745, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.543569564819336, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8498288989067078, "num_tokens": 118821456.0, "step": 3112 }, { "epoch": 0.396005597252258, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.565442085266113, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8538025617599487, "num_tokens": 118861746.0, "step": 3113 }, { "epoch": 0.3961328075308485, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.592284202575684, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8613158464431763, "num_tokens": 118900346.0, "step": 3114 }, { "epoch": 0.396260017809439, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.580289840698242, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8562674522399902, "num_tokens": 118939707.0, "step": 3115 }, { "epoch": 0.3963872280880295, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.56140422821045, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8554617166519165, "num_tokens": 118971154.0, "step": 3116 }, { "epoch": 0.39651443836662004, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.565348625183105, "learning_rate": 1e-06, "loss": 0.5062, "mean_token_accuracy": 0.8400325775146484, "num_tokens": 119012344.0, "step": 3117 }, { "epoch": 0.3966416486452105, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.540670394897461, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8548403382301331, "num_tokens": 119047677.0, "step": 3118 }, { "epoch": 0.39676885892380104, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.541498184204102, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8701890707015991, "num_tokens": 119086357.0, "step": 3119 }, { "epoch": 0.39689606920239157, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.538267135620117, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8626184463500977, "num_tokens": 119122854.0, "step": 3120 }, { "epoch": 0.39702327948098204, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.52826976776123, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8622640371322632, "num_tokens": 119155449.0, "step": 3121 }, { "epoch": 0.39715048975957257, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.547630310058594, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8508342504501343, "num_tokens": 119197287.0, "step": 3122 }, { "epoch": 0.3972777000381631, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.540411949157715, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8439421653747559, "num_tokens": 119235749.0, "step": 3123 }, { "epoch": 0.39740491031675357, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.539523124694824, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8516913652420044, "num_tokens": 119267282.0, "step": 3124 }, { "epoch": 0.3975321205953441, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.557342529296875, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8465766906738281, "num_tokens": 119304387.0, "step": 3125 }, { "epoch": 0.3976593308739346, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.531977653503418, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8459007740020752, "num_tokens": 119341470.0, "step": 3126 }, { "epoch": 0.3977865411525251, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.555649757385254, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8439304232597351, "num_tokens": 119381067.0, "step": 3127 }, { "epoch": 0.3979137514311156, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.526228904724121, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8558157682418823, "num_tokens": 119415680.0, "step": 3128 }, { "epoch": 0.39804096170970615, "ewc_loss": 0.01513671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 9.563520431518555, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8457031846046448, "num_tokens": 119454300.0, "step": 3129 }, { "epoch": 0.3981681719882967, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.582479476928711, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8679267168045044, "num_tokens": 119485599.0, "step": 3130 }, { "epoch": 0.39829538226688715, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.546270370483398, "learning_rate": 1e-06, "loss": 0.4928, "mean_token_accuracy": 0.8407514095306396, "num_tokens": 119525820.0, "step": 3131 }, { "epoch": 0.3984225925454777, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.529696464538574, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8608746528625488, "num_tokens": 119568216.0, "step": 3132 }, { "epoch": 0.3985498028240682, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.591547966003418, "learning_rate": 1e-06, "loss": 0.4848, "mean_token_accuracy": 0.8421995043754578, "num_tokens": 119602914.0, "step": 3133 }, { "epoch": 0.3986770131026587, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.562146186828613, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8517464399337769, "num_tokens": 119637024.0, "step": 3134 }, { "epoch": 0.3988042233812492, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.529271125793457, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.8457759618759155, "num_tokens": 119679265.0, "step": 3135 }, { "epoch": 0.39893143365983974, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.571296691894531, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8450593948364258, "num_tokens": 119716254.0, "step": 3136 }, { "epoch": 0.3990586439384302, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.522929191589355, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8474314212799072, "num_tokens": 119761325.0, "step": 3137 }, { "epoch": 0.39918585421702074, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.57177448272705, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8780257701873779, "num_tokens": 119795468.0, "step": 3138 }, { "epoch": 0.39931306449561127, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.549484252929688, "learning_rate": 1e-06, "loss": 0.4927, "mean_token_accuracy": 0.845413863658905, "num_tokens": 119833804.0, "step": 3139 }, { "epoch": 0.39944027477420174, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.547069549560547, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8498204350471497, "num_tokens": 119870852.0, "step": 3140 }, { "epoch": 0.39956748505279227, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.523335456848145, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8578637838363647, "num_tokens": 119909258.0, "step": 3141 }, { "epoch": 0.3996946953313828, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.57857608795166, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.8439245223999023, "num_tokens": 119940701.0, "step": 3142 }, { "epoch": 0.39982190560997327, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.54812240600586, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8554089665412903, "num_tokens": 119976780.0, "step": 3143 }, { "epoch": 0.3999491158885638, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.480877876281738, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8465062975883484, "num_tokens": 120017543.0, "step": 3144 }, { "epoch": 0.40007632616715433, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.583648681640625, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.851567268371582, "num_tokens": 120050036.0, "step": 3145 }, { "epoch": 0.4002035364457448, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.536911964416504, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8692641258239746, "num_tokens": 120084582.0, "step": 3146 }, { "epoch": 0.40033074672433533, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.525092124938965, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8506235480308533, "num_tokens": 120120218.0, "step": 3147 }, { "epoch": 0.40045795700292586, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.517571449279785, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8761835098266602, "num_tokens": 120161650.0, "step": 3148 }, { "epoch": 0.40058516728151633, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.523422241210938, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.864025354385376, "num_tokens": 120197409.0, "step": 3149 }, { "epoch": 0.40071237756010686, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.607213020324707, "learning_rate": 1e-06, "loss": 0.4923, "mean_token_accuracy": 0.8427964448928833, "num_tokens": 120240700.0, "step": 3150 }, { "epoch": 0.4008395878386974, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.502509117126465, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8456711173057556, "num_tokens": 120286741.0, "step": 3151 }, { "epoch": 0.40096679811728786, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.688170433044434, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.857305645942688, "num_tokens": 120323236.0, "step": 3152 }, { "epoch": 0.4010940083958784, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.546148300170898, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8539700508117676, "num_tokens": 120362185.0, "step": 3153 }, { "epoch": 0.4012212186744689, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.75667953491211, "learning_rate": 1e-06, "loss": 0.4736, "mean_token_accuracy": 0.8463824391365051, "num_tokens": 120397524.0, "step": 3154 }, { "epoch": 0.4013484289530594, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.520526885986328, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8584179878234863, "num_tokens": 120439723.0, "step": 3155 }, { "epoch": 0.4014756392316499, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.517950057983398, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8585180044174194, "num_tokens": 120483693.0, "step": 3156 }, { "epoch": 0.40160284951024044, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.639445304870605, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8648831844329834, "num_tokens": 120520558.0, "step": 3157 }, { "epoch": 0.4017300597888309, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.611823081970215, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8642572164535522, "num_tokens": 120557901.0, "step": 3158 }, { "epoch": 0.40185727006742145, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.551413536071777, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8607685565948486, "num_tokens": 120593851.0, "step": 3159 }, { "epoch": 0.401984480346012, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.59404468536377, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8469138145446777, "num_tokens": 120624533.0, "step": 3160 }, { "epoch": 0.40211169062460245, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.616384506225586, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8503601551055908, "num_tokens": 120660820.0, "step": 3161 }, { "epoch": 0.402238900903193, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 3.170967102050781e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 9.573995590209961, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8669186234474182, "num_tokens": 120701485.0, "step": 3162 }, { "epoch": 0.4023661111817835, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.596110343933105, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8645884990692139, "num_tokens": 120736677.0, "step": 3163 }, { "epoch": 0.402493321460374, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.560968399047852, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8529918193817139, "num_tokens": 120773933.0, "step": 3164 }, { "epoch": 0.4026205317389645, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.594029426574707, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8604074120521545, "num_tokens": 120804671.0, "step": 3165 }, { "epoch": 0.40274774201755503, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.591513633728027, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8599518537521362, "num_tokens": 120837042.0, "step": 3166 }, { "epoch": 0.4028749522961455, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.566479682922363, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8514391183853149, "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": 9.573166847229004, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8538339138031006, "num_tokens": 120915937.0, "step": 3168 }, { "epoch": 0.40312937285332656, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.544004440307617, "learning_rate": 1e-06, "loss": 0.4847, "mean_token_accuracy": 0.8398451805114746, "num_tokens": 120953334.0, "step": 3169 }, { "epoch": 0.40325658313191703, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.61569595336914, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8549684286117554, "num_tokens": 120986674.0, "step": 3170 }, { "epoch": 0.40338379341050756, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.56773567199707, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8617576360702515, "num_tokens": 121026305.0, "step": 3171 }, { "epoch": 0.4035110036890981, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.582058906555176, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8690147399902344, "num_tokens": 121061644.0, "step": 3172 }, { "epoch": 0.40363821396768856, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.563982963562012, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.851483941078186, "num_tokens": 121102860.0, "step": 3173 }, { "epoch": 0.4037654242462791, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.578642845153809, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8691371083259583, "num_tokens": 121136566.0, "step": 3174 }, { "epoch": 0.4038926345248696, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.613554000854492, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8653110861778259, "num_tokens": 121173759.0, "step": 3175 }, { "epoch": 0.4040198448034601, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.535179138183594, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8558520078659058, "num_tokens": 121219881.0, "step": 3176 }, { "epoch": 0.4041470550820506, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.600586891174316, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8541113138198853, "num_tokens": 121258719.0, "step": 3177 }, { "epoch": 0.40427426536064115, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.588674545288086, "learning_rate": 1e-06, "loss": 0.5131, "mean_token_accuracy": 0.8322319388389587, "num_tokens": 121298013.0, "step": 3178 }, { "epoch": 0.4044014756392316, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.603620529174805, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.866990327835083, "num_tokens": 121329615.0, "step": 3179 }, { "epoch": 0.40452868591782215, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.558467864990234, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8643008470535278, "num_tokens": 121365199.0, "step": 3180 }, { "epoch": 0.4046558961964127, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.520888328552246, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8511931896209717, "num_tokens": 121402393.0, "step": 3181 }, { "epoch": 0.4047831064750032, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.56222152709961, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8554797172546387, "num_tokens": 121439678.0, "step": 3182 }, { "epoch": 0.4049103167535937, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.543695449829102, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8498372435569763, "num_tokens": 121480781.0, "step": 3183 }, { "epoch": 0.4050375270321842, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.560266494750977, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8628474473953247, "num_tokens": 121517614.0, "step": 3184 }, { "epoch": 0.40516473731077474, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.557928085327148, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8531731963157654, "num_tokens": 121552225.0, "step": 3185 }, { "epoch": 0.4052919475893652, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.57814884185791, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8499994277954102, "num_tokens": 121587177.0, "step": 3186 }, { "epoch": 0.40541915786795574, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.615962982177734, "learning_rate": 1e-06, "loss": 0.4991, "mean_token_accuracy": 0.8375387191772461, "num_tokens": 121619547.0, "step": 3187 }, { "epoch": 0.40554636814654627, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.57278823852539, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8627277612686157, "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": 9.614577293395996, "learning_rate": 1e-06, "loss": 0.4924, "mean_token_accuracy": 0.8388722538948059, "num_tokens": 121691867.0, "step": 3189 }, { "epoch": 0.40580078870372727, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.60334300994873, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.849055290222168, "num_tokens": 121723546.0, "step": 3190 }, { "epoch": 0.4059279989823178, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.524124145507812, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8549219369888306, "num_tokens": 121764589.0, "step": 3191 }, { "epoch": 0.40605520926090827, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.600927352905273, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8442518711090088, "num_tokens": 121803281.0, "step": 3192 }, { "epoch": 0.4061824195394988, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.62048625946045, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8425730466842651, "num_tokens": 121841710.0, "step": 3193 }, { "epoch": 0.4063096298180893, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.598602294921875, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8529770970344543, "num_tokens": 121881979.0, "step": 3194 }, { "epoch": 0.4064368400966798, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.533334732055664, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8603452444076538, "num_tokens": 121921599.0, "step": 3195 }, { "epoch": 0.4065640503752703, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.632247924804688, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.863837480545044, "num_tokens": 121960545.0, "step": 3196 }, { "epoch": 0.40669126065386085, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.552576065063477, "learning_rate": 1e-06, "loss": 0.5078, "mean_token_accuracy": 0.8367162942886353, "num_tokens": 121999259.0, "step": 3197 }, { "epoch": 0.4068184709324513, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.599710464477539, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8681374788284302, "num_tokens": 122039066.0, "step": 3198 }, { "epoch": 0.40694568121104185, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.517630577087402, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8624826669692993, "num_tokens": 122085035.0, "step": 3199 }, { "epoch": 0.4070728914896324, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.56482219696045, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8695063591003418, "num_tokens": 122128007.0, "step": 3200 }, { "epoch": 0.40720010176822286, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.622912406921387, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8533077239990234, "num_tokens": 122169623.0, "step": 3201 }, { "epoch": 0.4073273120468134, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.675116539001465, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8509058356285095, "num_tokens": 122201755.0, "step": 3202 }, { "epoch": 0.4074545223254039, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.625462532043457, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8656229972839355, "num_tokens": 122233680.0, "step": 3203 }, { "epoch": 0.4075817326039944, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.677347183227539, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8522210121154785, "num_tokens": 122265107.0, "step": 3204 }, { "epoch": 0.4077089428825849, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.561073303222656, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8596168160438538, "num_tokens": 122302750.0, "step": 3205 }, { "epoch": 0.40783615316117544, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.582684516906738, "learning_rate": 1e-06, "loss": 0.4781, "mean_token_accuracy": 0.8468008041381836, "num_tokens": 122346820.0, "step": 3206 }, { "epoch": 0.4079633634397659, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.558348655700684, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8601783514022827, "num_tokens": 122383254.0, "step": 3207 }, { "epoch": 0.40809057371835644, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.654762268066406, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8673327565193176, "num_tokens": 122416693.0, "step": 3208 }, { "epoch": 0.40821778399694697, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.577677726745605, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8609087467193604, "num_tokens": 122452403.0, "step": 3209 }, { "epoch": 0.40834499427553744, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.634603500366211, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.844831109046936, "num_tokens": 122489685.0, "step": 3210 }, { "epoch": 0.40847220455412797, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.596179008483887, "learning_rate": 1e-06, "loss": 0.5005, "mean_token_accuracy": 0.8381695747375488, "num_tokens": 122529079.0, "step": 3211 }, { "epoch": 0.4085994148327185, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.677112579345703, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8487710952758789, "num_tokens": 122569933.0, "step": 3212 }, { "epoch": 0.40872662511130897, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.668217658996582, "learning_rate": 1e-06, "loss": 0.5085, "mean_token_accuracy": 0.835779070854187, "num_tokens": 122607884.0, "step": 3213 }, { "epoch": 0.4088538353898995, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.614877700805664, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.857428789138794, "num_tokens": 122651968.0, "step": 3214 }, { "epoch": 0.40898104566849003, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.617395401000977, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8595476746559143, "num_tokens": 122688680.0, "step": 3215 }, { "epoch": 0.4091082559470805, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.655510902404785, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8616117238998413, "num_tokens": 122728271.0, "step": 3216 }, { "epoch": 0.40923546622567103, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.622245788574219, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8547385334968567, "num_tokens": 122759691.0, "step": 3217 }, { "epoch": 0.40936267650426156, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.678733825683594, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8541927933692932, "num_tokens": 122800185.0, "step": 3218 }, { "epoch": 0.40948988678285203, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.551453590393066, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8647656440734863, "num_tokens": 122839548.0, "step": 3219 }, { "epoch": 0.40961709706144256, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.624217987060547, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8423984050750732, "num_tokens": 122879621.0, "step": 3220 }, { "epoch": 0.4097443073400331, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.63843822479248, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8509202599525452, "num_tokens": 122921495.0, "step": 3221 }, { "epoch": 0.40987151761862356, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.604867935180664, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8619878888130188, "num_tokens": 122957197.0, "step": 3222 }, { "epoch": 0.4099987278972141, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.609108924865723, "learning_rate": 1e-06, "loss": 0.5016, "mean_token_accuracy": 0.8373064994812012, "num_tokens": 122999999.0, "step": 3223 }, { "epoch": 0.4101259381758046, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.645414352416992, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8605141639709473, "num_tokens": 123040402.0, "step": 3224 }, { "epoch": 0.4102531484543951, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.571144104003906, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8598360419273376, "num_tokens": 123080086.0, "step": 3225 }, { "epoch": 0.4103803587329856, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.58116340637207, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8573430180549622, "num_tokens": 123120058.0, "step": 3226 }, { "epoch": 0.41050756901157615, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.575366973876953, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8568838834762573, "num_tokens": 123162003.0, "step": 3227 }, { "epoch": 0.4106347792901666, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.680191993713379, "learning_rate": 1e-06, "loss": 0.5413, "mean_token_accuracy": 0.8229537010192871, "num_tokens": 123202249.0, "step": 3228 }, { "epoch": 0.41076198956875715, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.588488578796387, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8495866060256958, "num_tokens": 123237401.0, "step": 3229 }, { "epoch": 0.4108891998473477, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.623594284057617, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8723502159118652, "num_tokens": 123269364.0, "step": 3230 }, { "epoch": 0.4110164101259382, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.607758522033691, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8582549691200256, "num_tokens": 123307743.0, "step": 3231 }, { "epoch": 0.4111436204045287, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 9.640764236450195, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8602402806282043, "num_tokens": 123340102.0, "step": 3232 }, { "epoch": 0.4112708306831192, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.649712562561035, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8508620262145996, "num_tokens": 123374956.0, "step": 3233 }, { "epoch": 0.41139804096170973, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.586079597473145, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8651561737060547, "num_tokens": 123418070.0, "step": 3234 }, { "epoch": 0.4115252512403002, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.623279571533203, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8529385924339294, "num_tokens": 123453183.0, "step": 3235 }, { "epoch": 0.41165246151889073, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.631028175354004, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8567372560501099, "num_tokens": 123491526.0, "step": 3236 }, { "epoch": 0.41177967179748126, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.606902122497559, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8579582571983337, "num_tokens": 123529743.0, "step": 3237 }, { "epoch": 0.41190688207607173, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.567723274230957, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8584351539611816, "num_tokens": 123573402.0, "step": 3238 }, { "epoch": 0.41203409235466226, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.683174133300781, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8530930280685425, "num_tokens": 123604167.0, "step": 3239 }, { "epoch": 0.4121613026332528, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.55133056640625, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8518234491348267, "num_tokens": 123648104.0, "step": 3240 }, { "epoch": 0.41228851291184326, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.587454795837402, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.845739483833313, "num_tokens": 123689678.0, "step": 3241 }, { "epoch": 0.4124157231904338, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.578287124633789, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8630490899085999, "num_tokens": 123729257.0, "step": 3242 }, { "epoch": 0.4125429334690243, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.6304292678833, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8579629063606262, "num_tokens": 123772251.0, "step": 3243 }, { "epoch": 0.4126701437476148, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.592960357666016, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8682852983474731, "num_tokens": 123810345.0, "step": 3244 }, { "epoch": 0.4127973540262053, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.628579139709473, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8522686958312988, "num_tokens": 123855276.0, "step": 3245 }, { "epoch": 0.41292456430479585, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.606002807617188, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.859017550945282, "num_tokens": 123896613.0, "step": 3246 }, { "epoch": 0.4130517745833863, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.660500526428223, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8399075269699097, "num_tokens": 123940713.0, "step": 3247 }, { "epoch": 0.41317898486197685, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.598718643188477, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8454631567001343, "num_tokens": 123981485.0, "step": 3248 }, { "epoch": 0.4133061951405674, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.671967506408691, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.856354832649231, "num_tokens": 124018159.0, "step": 3249 }, { "epoch": 0.41343340541915785, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 9.642839431762695, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8658881783485413, "num_tokens": 124049911.0, "step": 3250 }, { "epoch": 0.4135606156977484, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.626757621765137, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8767181634902954, "num_tokens": 124084260.0, "step": 3251 }, { "epoch": 0.4136878259763389, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.596766471862793, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8583204746246338, "num_tokens": 124122741.0, "step": 3252 }, { "epoch": 0.4138150362549294, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.645089149475098, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.852689802646637, "num_tokens": 124162793.0, "step": 3253 }, { "epoch": 0.4139422465335199, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.624469757080078, "learning_rate": 1e-06, "loss": 0.4693, "mean_token_accuracy": 0.849036693572998, "num_tokens": 124206889.0, "step": 3254 }, { "epoch": 0.41406945681211044, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.632674217224121, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8669419288635254, "num_tokens": 124244551.0, "step": 3255 }, { "epoch": 0.4141966670907009, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.617681503295898, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.858250081539154, "num_tokens": 124290485.0, "step": 3256 }, { "epoch": 0.41432387736929144, "ewc_loss": 0.015380859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.632530212402344, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8627347350120544, "num_tokens": 124323763.0, "step": 3257 }, { "epoch": 0.41445108764788197, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.653155326843262, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8558460474014282, "num_tokens": 124363318.0, "step": 3258 }, { "epoch": 0.41457829792647244, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.638338088989258, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8785163760185242, "num_tokens": 124400845.0, "step": 3259 }, { "epoch": 0.41470550820506297, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.64426040649414, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8618080615997314, "num_tokens": 124442546.0, "step": 3260 }, { "epoch": 0.4148327184836535, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.675590515136719, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8502070307731628, "num_tokens": 124477737.0, "step": 3261 }, { "epoch": 0.41495992876224397, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.647625923156738, "learning_rate": 1e-06, "loss": 0.4862, "mean_token_accuracy": 0.8430951833724976, "num_tokens": 124518081.0, "step": 3262 }, { "epoch": 0.4150871390408345, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.622580528259277, "learning_rate": 1e-06, "loss": 0.4736, "mean_token_accuracy": 0.8456553220748901, "num_tokens": 124556361.0, "step": 3263 }, { "epoch": 0.415214349319425, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.669326782226562, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8549468517303467, "num_tokens": 124588490.0, "step": 3264 }, { "epoch": 0.4153415595980155, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.614131927490234, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8540709018707275, "num_tokens": 124624656.0, "step": 3265 }, { "epoch": 0.415468769876606, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.660318374633789, "learning_rate": 1e-06, "loss": 0.4906, "mean_token_accuracy": 0.8428694605827332, "num_tokens": 124660681.0, "step": 3266 }, { "epoch": 0.41559598015519655, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.611581802368164, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8566247224807739, "num_tokens": 124703448.0, "step": 3267 }, { "epoch": 0.415723190433787, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.602669715881348, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8534941673278809, "num_tokens": 124742078.0, "step": 3268 }, { "epoch": 0.41585040071237755, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.671706199645996, "learning_rate": 1e-06, "loss": 0.4956, "mean_token_accuracy": 0.843393087387085, "num_tokens": 124779894.0, "step": 3269 }, { "epoch": 0.4159776109909681, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.632097244262695, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8504258990287781, "num_tokens": 124817918.0, "step": 3270 }, { "epoch": 0.41610482126955856, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.63072395324707, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8718173503875732, "num_tokens": 124854839.0, "step": 3271 }, { "epoch": 0.4162320315481491, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.656664848327637, "learning_rate": 1e-06, "loss": 0.4946, "mean_token_accuracy": 0.8401869535446167, "num_tokens": 124886779.0, "step": 3272 }, { "epoch": 0.4163592418267396, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.618816375732422, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8637682795524597, "num_tokens": 124927919.0, "step": 3273 }, { "epoch": 0.4164864521053301, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.622614860534668, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8526893854141235, "num_tokens": 124966475.0, "step": 3274 }, { "epoch": 0.4166136623839206, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.6654052734375, "learning_rate": 1e-06, "loss": 0.5282, "mean_token_accuracy": 0.8312628269195557, "num_tokens": 125004110.0, "step": 3275 }, { "epoch": 0.41674087266251114, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.58515453338623, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8699541091918945, "num_tokens": 125043164.0, "step": 3276 }, { "epoch": 0.4168680829411016, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.668387413024902, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8584002256393433, "num_tokens": 125084255.0, "step": 3277 }, { "epoch": 0.41699529321969214, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.628548622131348, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8685427308082581, "num_tokens": 125124971.0, "step": 3278 }, { "epoch": 0.41712250349828267, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.6923828125, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8516092896461487, "num_tokens": 125161470.0, "step": 3279 }, { "epoch": 0.4172497137768732, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.671919822692871, "learning_rate": 1e-06, "loss": 0.4937, "mean_token_accuracy": 0.8403438925743103, "num_tokens": 125199117.0, "step": 3280 }, { "epoch": 0.41737692405546367, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.617510795593262, "learning_rate": 1e-06, "loss": 0.5066, "mean_token_accuracy": 0.8377109169960022, "num_tokens": 125238136.0, "step": 3281 }, { "epoch": 0.4175041343340542, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.608283996582031, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8533735275268555, "num_tokens": 125281041.0, "step": 3282 }, { "epoch": 0.41763134461264473, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.657772064208984, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8579119443893433, "num_tokens": 125314739.0, "step": 3283 }, { "epoch": 0.4177585548912352, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.670774459838867, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8493517637252808, "num_tokens": 125353129.0, "step": 3284 }, { "epoch": 0.41788576516982573, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.70751667022705, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8544792532920837, "num_tokens": 125394092.0, "step": 3285 }, { "epoch": 0.41801297544841626, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.670809745788574, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8494075536727905, "num_tokens": 125438262.0, "step": 3286 }, { "epoch": 0.41814018572700673, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.629088401794434, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8539784550666809, "num_tokens": 125477842.0, "step": 3287 }, { "epoch": 0.41826739600559726, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.631699562072754, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8672741651535034, "num_tokens": 125517236.0, "step": 3288 }, { "epoch": 0.4183946062841878, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.670877456665039, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8480645418167114, "num_tokens": 125561598.0, "step": 3289 }, { "epoch": 0.41852181656277826, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.634586334228516, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8535601496696472, "num_tokens": 125599104.0, "step": 3290 }, { "epoch": 0.4186490268413688, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.6534423828125, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8558803796768188, "num_tokens": 125635286.0, "step": 3291 }, { "epoch": 0.4187762371199593, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.697480201721191, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8598927855491638, "num_tokens": 125669534.0, "step": 3292 }, { "epoch": 0.4189034473985498, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.626790046691895, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8547273874282837, "num_tokens": 125707966.0, "step": 3293 }, { "epoch": 0.4190306576771403, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.623342514038086, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8499829769134521, "num_tokens": 125741448.0, "step": 3294 }, { "epoch": 0.41915786795573085, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.678638458251953, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8484709858894348, "num_tokens": 125776887.0, "step": 3295 }, { "epoch": 0.4192850782343213, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.663833618164062, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.859369158744812, "num_tokens": 125807899.0, "step": 3296 }, { "epoch": 0.41941228851291185, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.639002799987793, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8635992407798767, "num_tokens": 125847711.0, "step": 3297 }, { "epoch": 0.4195394987915024, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.634927749633789, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8685822486877441, "num_tokens": 125883248.0, "step": 3298 }, { "epoch": 0.41966670907009285, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.644556045532227, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8433741331100464, "num_tokens": 125925771.0, "step": 3299 }, { "epoch": 0.4197939193486834, "ewc_loss": 0.015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.642837524414062, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8434630632400513, "num_tokens": 125968815.0, "step": 3300 }, { "epoch": 0.4199211296272739, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.639034271240234, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8511251211166382, "num_tokens": 126007761.0, "step": 3301 }, { "epoch": 0.4200483399058644, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.677897453308105, "learning_rate": 1e-06, "loss": 0.5025, "mean_token_accuracy": 0.8360021114349365, "num_tokens": 126047872.0, "step": 3302 }, { "epoch": 0.4201755501844549, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.62896728515625, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8623079061508179, "num_tokens": 126081273.0, "step": 3303 }, { "epoch": 0.42030276046304543, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.675515174865723, "learning_rate": 1e-06, "loss": 0.4975, "mean_token_accuracy": 0.8402790427207947, "num_tokens": 126121590.0, "step": 3304 }, { "epoch": 0.4204299707416359, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.698726654052734, "learning_rate": 1e-06, "loss": 0.4922, "mean_token_accuracy": 0.8382784128189087, "num_tokens": 126168396.0, "step": 3305 }, { "epoch": 0.42055718102022643, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 9.631623268127441, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8503628969192505, "num_tokens": 126205220.0, "step": 3306 }, { "epoch": 0.42068439129881696, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.74255084991455, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8484823703765869, "num_tokens": 126239131.0, "step": 3307 }, { "epoch": 0.42081160157740743, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.679230690002441, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8571388721466064, "num_tokens": 126271931.0, "step": 3308 }, { "epoch": 0.42093881185599796, "ewc_loss": 0.015625, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.671363830566406, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8638161420822144, "num_tokens": 126307730.0, "step": 3309 }, { "epoch": 0.4210660221345885, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.68946361541748, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8587210178375244, "num_tokens": 126344528.0, "step": 3310 }, { "epoch": 0.42119323241317896, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.696918487548828, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8641757965087891, "num_tokens": 126377260.0, "step": 3311 }, { "epoch": 0.4213204426917695, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.687906265258789, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8511480093002319, "num_tokens": 126409723.0, "step": 3312 }, { "epoch": 0.42144765297036, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.67648696899414, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.85878586769104, "num_tokens": 126450929.0, "step": 3313 }, { "epoch": 0.4215748632489505, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.716036796569824, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8648878931999207, "num_tokens": 126486878.0, "step": 3314 }, { "epoch": 0.421702073527541, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.679640769958496, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8432697653770447, "num_tokens": 126522884.0, "step": 3315 }, { "epoch": 0.42182928380613155, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.736163139343262, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8560085296630859, "num_tokens": 126560848.0, "step": 3316 }, { "epoch": 0.421956494084722, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.639541625976562, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.857302725315094, "num_tokens": 126592947.0, "step": 3317 }, { "epoch": 0.42208370436331255, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.693479537963867, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8658849596977234, "num_tokens": 126630631.0, "step": 3318 }, { "epoch": 0.4222109146419031, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.684081077575684, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8512209057807922, "num_tokens": 126673395.0, "step": 3319 }, { "epoch": 0.42233812492049355, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.666675567626953, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8712282180786133, "num_tokens": 126711504.0, "step": 3320 }, { "epoch": 0.4224653351990841, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.67685604095459, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8686866164207458, "num_tokens": 126753217.0, "step": 3321 }, { "epoch": 0.4225925454776746, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.683066368103027, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8611057996749878, "num_tokens": 126792982.0, "step": 3322 }, { "epoch": 0.4227197557562651, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.732608795166016, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8495208621025085, "num_tokens": 126823771.0, "step": 3323 }, { "epoch": 0.4228469660348556, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.655238151550293, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8595972061157227, "num_tokens": 126867137.0, "step": 3324 }, { "epoch": 0.42297417631344614, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.630600929260254, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8610935807228088, "num_tokens": 126902680.0, "step": 3325 }, { "epoch": 0.4231013865920366, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.71206283569336, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.866768479347229, "num_tokens": 126941037.0, "step": 3326 }, { "epoch": 0.42322859687062714, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.64601993560791, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8596584796905518, "num_tokens": 126980009.0, "step": 3327 }, { "epoch": 0.42335580714921767, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.653482437133789, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8686708211898804, "num_tokens": 127020022.0, "step": 3328 }, { "epoch": 0.42348301742780814, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.70366096496582, "learning_rate": 1e-06, "loss": 0.5061, "mean_token_accuracy": 0.8360224366188049, "num_tokens": 127060500.0, "step": 3329 }, { "epoch": 0.42361022770639867, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.684995651245117, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8519489765167236, "num_tokens": 127095886.0, "step": 3330 }, { "epoch": 0.4237374379849892, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.676265716552734, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8464665412902832, "num_tokens": 127134334.0, "step": 3331 }, { "epoch": 0.4238646482635797, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.640280723571777, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8482882976531982, "num_tokens": 127174288.0, "step": 3332 }, { "epoch": 0.4239918585421702, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.665650367736816, "learning_rate": 1e-06, "loss": 0.4781, "mean_token_accuracy": 0.8533464074134827, "num_tokens": 127210311.0, "step": 3333 }, { "epoch": 0.4241190688207607, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.659008026123047, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8768606185913086, "num_tokens": 127252813.0, "step": 3334 }, { "epoch": 0.42424627909935125, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.725876808166504, "learning_rate": 1e-06, "loss": 0.4835, "mean_token_accuracy": 0.8438614010810852, "num_tokens": 127285847.0, "step": 3335 }, { "epoch": 0.4243734893779417, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.682002067565918, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8702661991119385, "num_tokens": 127325360.0, "step": 3336 }, { "epoch": 0.42450069965653225, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.6766357421875, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8595611453056335, "num_tokens": 127366446.0, "step": 3337 }, { "epoch": 0.4246279099351228, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.695391654968262, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.865073025226593, "num_tokens": 127402608.0, "step": 3338 }, { "epoch": 0.42475512021371326, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.7111177444458, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8621329069137573, "num_tokens": 127445355.0, "step": 3339 }, { "epoch": 0.4248823304923038, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.69177532196045, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8508254289627075, "num_tokens": 127480523.0, "step": 3340 }, { "epoch": 0.4250095407708943, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.680352210998535, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.87578946352005, "num_tokens": 127520787.0, "step": 3341 }, { "epoch": 0.4251367510494848, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.664958000183105, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.851591169834137, "num_tokens": 127565576.0, "step": 3342 }, { "epoch": 0.4252639613280753, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.747870445251465, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8672068119049072, "num_tokens": 127599144.0, "step": 3343 }, { "epoch": 0.42539117160666584, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.661681175231934, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8747045993804932, "num_tokens": 127637361.0, "step": 3344 }, { "epoch": 0.4255183818852563, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.736581802368164, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8460306525230408, "num_tokens": 127675139.0, "step": 3345 }, { "epoch": 0.42564559216384684, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.640443801879883, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8621261119842529, "num_tokens": 127718822.0, "step": 3346 }, { "epoch": 0.42577280244243737, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.701704978942871, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8600583076477051, "num_tokens": 127763332.0, "step": 3347 }, { "epoch": 0.42590001272102784, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.684968948364258, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8678430914878845, "num_tokens": 127800346.0, "step": 3348 }, { "epoch": 0.42602722299961837, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.69594669342041, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8537364602088928, "num_tokens": 127845142.0, "step": 3349 }, { "epoch": 0.4261544332782089, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.704354286193848, "learning_rate": 1e-06, "loss": 0.5082, "mean_token_accuracy": 0.8395301103591919, "num_tokens": 127881131.0, "step": 3350 }, { "epoch": 0.4262816435567994, "ewc_loss": 0.015625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.685505867004395, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8707113265991211, "num_tokens": 127917941.0, "step": 3351 }, { "epoch": 0.4264088538353899, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.700791358947754, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8437075018882751, "num_tokens": 127963345.0, "step": 3352 }, { "epoch": 0.42653606411398043, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 9.662482261657715, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.865132212638855, "num_tokens": 128004839.0, "step": 3353 }, { "epoch": 0.4266632743925709, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.714770317077637, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8494383096694946, "num_tokens": 128042709.0, "step": 3354 }, { "epoch": 0.42679048467116143, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.650959968566895, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.842269778251648, "num_tokens": 128085069.0, "step": 3355 }, { "epoch": 0.42691769494975196, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.688047409057617, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8509689569473267, "num_tokens": 128124300.0, "step": 3356 }, { "epoch": 0.42704490522834243, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.64414119720459, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8547811508178711, "num_tokens": 128164410.0, "step": 3357 }, { "epoch": 0.42717211550693296, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.669445991516113, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8663057088851929, "num_tokens": 128206524.0, "step": 3358 }, { "epoch": 0.4272993257855235, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.648931503295898, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8478769659996033, "num_tokens": 128246387.0, "step": 3359 }, { "epoch": 0.42742653606411396, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.7007474899292, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.866542398929596, "num_tokens": 128287794.0, "step": 3360 }, { "epoch": 0.4275537463427045, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.65451717376709, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8544367551803589, "num_tokens": 128328467.0, "step": 3361 }, { "epoch": 0.427680956621295, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.684380531311035, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8661131858825684, "num_tokens": 128364319.0, "step": 3362 }, { "epoch": 0.4278081668998855, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.694592475891113, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8522840738296509, "num_tokens": 128403892.0, "step": 3363 }, { "epoch": 0.427935377178476, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.754781723022461, "learning_rate": 1e-06, "loss": 0.5078, "mean_token_accuracy": 0.8368529081344604, "num_tokens": 128446760.0, "step": 3364 }, { "epoch": 0.42806258745706655, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.706218719482422, "learning_rate": 1e-06, "loss": 0.5194, "mean_token_accuracy": 0.8321784734725952, "num_tokens": 128485498.0, "step": 3365 }, { "epoch": 0.428189797735657, "ewc_loss": 0.015625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.736800193786621, "learning_rate": 1e-06, "loss": 0.4943, "mean_token_accuracy": 0.8453075885772705, "num_tokens": 128527871.0, "step": 3366 }, { "epoch": 0.42831700801424755, "ewc_loss": 0.015625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.640247344970703, "learning_rate": 1e-06, "loss": 0.4816, "mean_token_accuracy": 0.8429234027862549, "num_tokens": 128567792.0, "step": 3367 }, { "epoch": 0.4284442182928381, "ewc_loss": 0.015625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.720928192138672, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8463552594184875, "num_tokens": 128607217.0, "step": 3368 }, { "epoch": 0.42857142857142855, "ewc_loss": 0.015625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 9.678838729858398, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8619714975357056, "num_tokens": 128642787.0, "step": 3369 }, { "epoch": 0.4286986388500191, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 9.683342933654785, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8670136332511902, "num_tokens": 128680462.0, "step": 3370 }, { "epoch": 0.4288258491286096, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.709501266479492, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.852882981300354, "num_tokens": 128715101.0, "step": 3371 }, { "epoch": 0.4289530594072001, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 9.742572784423828, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8479092121124268, "num_tokens": 128749601.0, "step": 3372 }, { "epoch": 0.4290802696857906, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.739273071289062, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.851293683052063, "num_tokens": 128788744.0, "step": 3373 }, { "epoch": 0.42920747996438113, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.72730827331543, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8517439365386963, "num_tokens": 128828102.0, "step": 3374 }, { "epoch": 0.4293346902429716, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.713323593139648, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8502360582351685, "num_tokens": 128863610.0, "step": 3375 }, { "epoch": 0.42946190052156213, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.692994117736816, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8698270916938782, "num_tokens": 128898836.0, "step": 3376 }, { "epoch": 0.42958911080015266, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.7101411819458, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8578450679779053, "num_tokens": 128934821.0, "step": 3377 }, { "epoch": 0.42971632107874314, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.696996688842773, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8538215756416321, "num_tokens": 128966914.0, "step": 3378 }, { "epoch": 0.42984353135733366, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.701436042785645, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8625325560569763, "num_tokens": 129008305.0, "step": 3379 }, { "epoch": 0.4299707416359242, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.668771743774414, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8761866688728333, "num_tokens": 129043331.0, "step": 3380 }, { "epoch": 0.4300979519145147, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.675472259521484, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8613277673721313, "num_tokens": 129083513.0, "step": 3381 }, { "epoch": 0.4302251621931052, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.70387077331543, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8548876047134399, "num_tokens": 129113463.0, "step": 3382 }, { "epoch": 0.4303523724716957, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.666756629943848, "learning_rate": 1e-06, "loss": 0.5078, "mean_token_accuracy": 0.8325167298316956, "num_tokens": 129153080.0, "step": 3383 }, { "epoch": 0.43047958275028625, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.721843719482422, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8526219129562378, "num_tokens": 129193338.0, "step": 3384 }, { "epoch": 0.4306067930288767, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.662096977233887, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8452139496803284, "num_tokens": 129234816.0, "step": 3385 }, { "epoch": 0.43073400330746725, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.670769691467285, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8587150573730469, "num_tokens": 129275228.0, "step": 3386 }, { "epoch": 0.4308612135860578, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.700249671936035, "learning_rate": 1e-06, "loss": 0.5024, "mean_token_accuracy": 0.8381518125534058, "num_tokens": 129309747.0, "step": 3387 }, { "epoch": 0.43098842386464825, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.730277061462402, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8592842817306519, "num_tokens": 129342580.0, "step": 3388 }, { "epoch": 0.4311156341432388, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.715954780578613, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8508889079093933, "num_tokens": 129376599.0, "step": 3389 }, { "epoch": 0.4312428444218293, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.686010360717773, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.854427695274353, "num_tokens": 129414783.0, "step": 3390 }, { "epoch": 0.4313700547004198, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.661927223205566, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8549920320510864, "num_tokens": 129450287.0, "step": 3391 }, { "epoch": 0.4314972649790103, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.76451587677002, "learning_rate": 1e-06, "loss": 0.5005, "mean_token_accuracy": 0.8368364572525024, "num_tokens": 129486733.0, "step": 3392 }, { "epoch": 0.43162447525760084, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.689048767089844, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8558932542800903, "num_tokens": 129524392.0, "step": 3393 }, { "epoch": 0.4317516855361913, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.679910659790039, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.842435359954834, "num_tokens": 129558452.0, "step": 3394 }, { "epoch": 0.43187889581478184, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.724115371704102, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8536175489425659, "num_tokens": 129597076.0, "step": 3395 }, { "epoch": 0.43200610609337237, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.695711135864258, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.849760115146637, "num_tokens": 129635861.0, "step": 3396 }, { "epoch": 0.43213331637196284, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.742096900939941, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8503351211547852, "num_tokens": 129673760.0, "step": 3397 }, { "epoch": 0.43226052665055337, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.665563583374023, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8566956520080566, "num_tokens": 129710730.0, "step": 3398 }, { "epoch": 0.4323877369291439, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.754227638244629, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8520535230636597, "num_tokens": 129748724.0, "step": 3399 }, { "epoch": 0.43251494720773437, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.75300121307373, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8682616353034973, "num_tokens": 129787478.0, "step": 3400 }, { "epoch": 0.4326421574863249, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.679460525512695, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8657188415527344, "num_tokens": 129829852.0, "step": 3401 }, { "epoch": 0.4327693677649154, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.704313278198242, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8550237417221069, "num_tokens": 129869279.0, "step": 3402 }, { "epoch": 0.4328965780435059, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.765970230102539, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8495816588401794, "num_tokens": 129903347.0, "step": 3403 }, { "epoch": 0.4330237883220964, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.70976734161377, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8657907247543335, "num_tokens": 129940028.0, "step": 3404 }, { "epoch": 0.43315099860068695, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.792291641235352, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8645086884498596, "num_tokens": 129976438.0, "step": 3405 }, { "epoch": 0.4332782088792774, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.675726890563965, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8606589436531067, "num_tokens": 130015698.0, "step": 3406 }, { "epoch": 0.43340541915786795, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.784293174743652, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8701034784317017, "num_tokens": 130056000.0, "step": 3407 }, { "epoch": 0.4335326294364585, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.674320220947266, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8530781269073486, "num_tokens": 130096990.0, "step": 3408 }, { "epoch": 0.43365983971504896, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.769826889038086, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8657292127609253, "num_tokens": 130128556.0, "step": 3409 }, { "epoch": 0.4337870499936395, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.686185836791992, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8618931770324707, "num_tokens": 130170607.0, "step": 3410 }, { "epoch": 0.43391426027223, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.724791526794434, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8519614934921265, "num_tokens": 130212511.0, "step": 3411 }, { "epoch": 0.4340414705508205, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.761397361755371, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8491743803024292, "num_tokens": 130249246.0, "step": 3412 }, { "epoch": 0.434168680829411, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.663722038269043, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8762496113777161, "num_tokens": 130291250.0, "step": 3413 }, { "epoch": 0.43429589110800154, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.747847557067871, "learning_rate": 1e-06, "loss": 0.503, "mean_token_accuracy": 0.8362636566162109, "num_tokens": 130327094.0, "step": 3414 }, { "epoch": 0.434423101386592, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.767108917236328, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8477027416229248, "num_tokens": 130366482.0, "step": 3415 }, { "epoch": 0.43455031166518254, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.721301078796387, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8556953072547913, "num_tokens": 130407416.0, "step": 3416 }, { "epoch": 0.43467752194377307, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.687590599060059, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8619237542152405, "num_tokens": 130447125.0, "step": 3417 }, { "epoch": 0.43480473222236354, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.75770092010498, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8616903424263, "num_tokens": 130480955.0, "step": 3418 }, { "epoch": 0.43493194250095407, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.75411319732666, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8682644963264465, "num_tokens": 130511771.0, "step": 3419 }, { "epoch": 0.4350591527795446, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.708803176879883, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8544061183929443, "num_tokens": 130550080.0, "step": 3420 }, { "epoch": 0.4351863630581351, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.709118843078613, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8440936803817749, "num_tokens": 130588109.0, "step": 3421 }, { "epoch": 0.4353135733367256, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.773036003112793, "learning_rate": 1e-06, "loss": 0.5015, "mean_token_accuracy": 0.8417137265205383, "num_tokens": 130626832.0, "step": 3422 }, { "epoch": 0.43544078361531613, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.710719108581543, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8604153394699097, "num_tokens": 130662202.0, "step": 3423 }, { "epoch": 0.4355679938939066, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.770132064819336, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8511301875114441, "num_tokens": 130700747.0, "step": 3424 }, { "epoch": 0.43569520417249713, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.71274185180664, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8542914390563965, "num_tokens": 130742849.0, "step": 3425 }, { "epoch": 0.43582241445108766, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.741697311401367, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8692501783370972, "num_tokens": 130780156.0, "step": 3426 }, { "epoch": 0.43594962472967813, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.7003812789917, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8559736013412476, "num_tokens": 130825824.0, "step": 3427 }, { "epoch": 0.43607683500826866, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.748510360717773, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8649561405181885, "num_tokens": 130863603.0, "step": 3428 }, { "epoch": 0.4362040452868592, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.707945823669434, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8579967021942139, "num_tokens": 130906689.0, "step": 3429 }, { "epoch": 0.4363312555654497, "ewc_loss": 0.015869140625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 9.747360229492188, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8535290360450745, "num_tokens": 130947785.0, "step": 3430 }, { "epoch": 0.4364584658440402, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.722862243652344, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.866277277469635, "num_tokens": 130985489.0, "step": 3431 }, { "epoch": 0.4365856761226307, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.72254467010498, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8646506071090698, "num_tokens": 131025274.0, "step": 3432 }, { "epoch": 0.43671288640122125, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.733414649963379, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.855755090713501, "num_tokens": 131060132.0, "step": 3433 }, { "epoch": 0.4368400966798117, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.726847648620605, "learning_rate": 1e-06, "loss": 0.4861, "mean_token_accuracy": 0.8452631235122681, "num_tokens": 131105109.0, "step": 3434 }, { "epoch": 0.43696730695840225, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.775696754455566, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8715578317642212, "num_tokens": 131142094.0, "step": 3435 }, { "epoch": 0.4370945172369928, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.770264625549316, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8542683124542236, "num_tokens": 131174648.0, "step": 3436 }, { "epoch": 0.43722172751558325, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.702391624450684, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8716804385185242, "num_tokens": 131213447.0, "step": 3437 }, { "epoch": 0.4373489377941738, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.78525161743164, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8518540859222412, "num_tokens": 131249622.0, "step": 3438 }, { "epoch": 0.4374761480727643, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.734976768493652, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8724861145019531, "num_tokens": 131285883.0, "step": 3439 }, { "epoch": 0.4376033583513548, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.70511531829834, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8597175478935242, "num_tokens": 131322774.0, "step": 3440 }, { "epoch": 0.4377305686299453, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 9.739045143127441, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8685767650604248, "num_tokens": 131364513.0, "step": 3441 }, { "epoch": 0.43785777890853583, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.736937522888184, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8475050926208496, "num_tokens": 131402299.0, "step": 3442 }, { "epoch": 0.4379849891871263, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.748360633850098, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8637325763702393, "num_tokens": 131447830.0, "step": 3443 }, { "epoch": 0.43811219946571683, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.781013488769531, "learning_rate": 1e-06, "loss": 0.4719, "mean_token_accuracy": 0.8497070074081421, "num_tokens": 131481058.0, "step": 3444 }, { "epoch": 0.43823940974430736, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.76028060913086, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8527832627296448, "num_tokens": 131518376.0, "step": 3445 }, { "epoch": 0.43836662002289783, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.739310264587402, "learning_rate": 1e-06, "loss": 0.5063, "mean_token_accuracy": 0.8425034284591675, "num_tokens": 131553702.0, "step": 3446 }, { "epoch": 0.43849383030148836, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.740880012512207, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8651092648506165, "num_tokens": 131587026.0, "step": 3447 }, { "epoch": 0.4386210405800789, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.709612846374512, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8627154231071472, "num_tokens": 131621145.0, "step": 3448 }, { "epoch": 0.43874825085866936, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.765273094177246, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8623943328857422, "num_tokens": 131661618.0, "step": 3449 }, { "epoch": 0.4388754611372599, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.746257781982422, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8513273596763611, "num_tokens": 131693840.0, "step": 3450 }, { "epoch": 0.4390026714158504, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.720105171203613, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8595313429832458, "num_tokens": 131730856.0, "step": 3451 }, { "epoch": 0.4391298816944409, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.79446792602539, "learning_rate": 1e-06, "loss": 0.5165, "mean_token_accuracy": 0.8334379196166992, "num_tokens": 131769172.0, "step": 3452 }, { "epoch": 0.4392570919730314, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.705123901367188, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8594958782196045, "num_tokens": 131810481.0, "step": 3453 }, { "epoch": 0.43938430225162195, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.660588264465332, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8586869239807129, "num_tokens": 131850742.0, "step": 3454 }, { "epoch": 0.4395115125302124, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.799253463745117, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8556227087974548, "num_tokens": 131886675.0, "step": 3455 }, { "epoch": 0.43963872280880295, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.708829879760742, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8629312515258789, "num_tokens": 131926831.0, "step": 3456 }, { "epoch": 0.4397659330873935, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.773048400878906, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.856169581413269, "num_tokens": 131962265.0, "step": 3457 }, { "epoch": 0.43989314336598395, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.7618408203125, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8525823950767517, "num_tokens": 131999906.0, "step": 3458 }, { "epoch": 0.4400203536445745, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.755809783935547, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8738710880279541, "num_tokens": 132031330.0, "step": 3459 }, { "epoch": 0.440147563923165, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.778629302978516, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8556171655654907, "num_tokens": 132065049.0, "step": 3460 }, { "epoch": 0.4402747742017555, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.746031761169434, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8616527318954468, "num_tokens": 132100993.0, "step": 3461 }, { "epoch": 0.440401984480346, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.73038387298584, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8553047776222229, "num_tokens": 132138169.0, "step": 3462 }, { "epoch": 0.44052919475893654, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.770790100097656, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8639751672744751, "num_tokens": 132170498.0, "step": 3463 }, { "epoch": 0.440656405037527, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.743963241577148, "learning_rate": 1e-06, "loss": 0.5012, "mean_token_accuracy": 0.8377740383148193, "num_tokens": 132213583.0, "step": 3464 }, { "epoch": 0.44078361531611754, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.701055526733398, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8519437313079834, "num_tokens": 132255427.0, "step": 3465 }, { "epoch": 0.44091082559470807, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.792778968811035, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8552918434143066, "num_tokens": 132295795.0, "step": 3466 }, { "epoch": 0.44103803587329854, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.7874755859375, "learning_rate": 1e-06, "loss": 0.4929, "mean_token_accuracy": 0.8390672206878662, "num_tokens": 132336881.0, "step": 3467 }, { "epoch": 0.44116524615188907, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.705853462219238, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.856484591960907, "num_tokens": 132375136.0, "step": 3468 }, { "epoch": 0.4412924564304796, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.801459312438965, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8542853593826294, "num_tokens": 132412893.0, "step": 3469 }, { "epoch": 0.44141966670907007, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.728797912597656, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.851134717464447, "num_tokens": 132444360.0, "step": 3470 }, { "epoch": 0.4415468769876606, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.726469039916992, "learning_rate": 1e-06, "loss": 0.5396, "mean_token_accuracy": 0.8308312892913818, "num_tokens": 132485022.0, "step": 3471 }, { "epoch": 0.4416740872662511, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.764300346374512, "learning_rate": 1e-06, "loss": 0.5108, "mean_token_accuracy": 0.8415968418121338, "num_tokens": 132522800.0, "step": 3472 }, { "epoch": 0.4418012975448416, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.763118743896484, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8591460585594177, "num_tokens": 132560527.0, "step": 3473 }, { "epoch": 0.4419285078234321, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.759145736694336, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8538250923156738, "num_tokens": 132601075.0, "step": 3474 }, { "epoch": 0.44205571810202265, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.732065200805664, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8474011421203613, "num_tokens": 132639851.0, "step": 3475 }, { "epoch": 0.4421829283806131, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.763814926147461, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8584767580032349, "num_tokens": 132679849.0, "step": 3476 }, { "epoch": 0.44231013865920366, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.733512878417969, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8616442680358887, "num_tokens": 132721020.0, "step": 3477 }, { "epoch": 0.4424373489377942, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.743842124938965, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8434028625488281, "num_tokens": 132760933.0, "step": 3478 }, { "epoch": 0.44256455921638466, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.736885070800781, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8676470518112183, "num_tokens": 132800627.0, "step": 3479 }, { "epoch": 0.4426917694949752, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.785096168518066, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8555940389633179, "num_tokens": 132833232.0, "step": 3480 }, { "epoch": 0.4428189797735657, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.703879356384277, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8544749617576599, "num_tokens": 132874670.0, "step": 3481 }, { "epoch": 0.44294619005215624, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.781163215637207, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8588076829910278, "num_tokens": 132916187.0, "step": 3482 }, { "epoch": 0.4430734003307467, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.745370864868164, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8594244122505188, "num_tokens": 132956948.0, "step": 3483 }, { "epoch": 0.44320061060933724, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.76888370513916, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8616272211074829, "num_tokens": 132995213.0, "step": 3484 }, { "epoch": 0.44332782088792777, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.759456634521484, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8532785773277283, "num_tokens": 133038850.0, "step": 3485 }, { "epoch": 0.44345503116651824, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.764430046081543, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8580204248428345, "num_tokens": 133080766.0, "step": 3486 }, { "epoch": 0.44358224144510877, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.716039657592773, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8533766269683838, "num_tokens": 133120845.0, "step": 3487 }, { "epoch": 0.4437094517236993, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.763985633850098, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8534655570983887, "num_tokens": 133166011.0, "step": 3488 }, { "epoch": 0.4438366620022898, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.753007888793945, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8545926809310913, "num_tokens": 133203507.0, "step": 3489 }, { "epoch": 0.4439638722808803, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.802072525024414, "learning_rate": 1e-06, "loss": 0.5153, "mean_token_accuracy": 0.8329632878303528, "num_tokens": 133240554.0, "step": 3490 }, { "epoch": 0.44409108255947083, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.807774543762207, "learning_rate": 1e-06, "loss": 0.5026, "mean_token_accuracy": 0.8394240140914917, "num_tokens": 133274987.0, "step": 3491 }, { "epoch": 0.4442182928380613, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.747282028198242, "learning_rate": 1e-06, "loss": 0.526, "mean_token_accuracy": 0.8293842077255249, "num_tokens": 133319287.0, "step": 3492 }, { "epoch": 0.44434550311665183, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.768887519836426, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8606848120689392, "num_tokens": 133353670.0, "step": 3493 }, { "epoch": 0.44447271339524236, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 9.720929145812988, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8580106496810913, "num_tokens": 133391950.0, "step": 3494 }, { "epoch": 0.44459992367383283, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.784382820129395, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8614397048950195, "num_tokens": 133424758.0, "step": 3495 }, { "epoch": 0.44472713395242336, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.811361312866211, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8845396637916565, "num_tokens": 133459869.0, "step": 3496 }, { "epoch": 0.4448543442310139, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.779325485229492, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8549702763557434, "num_tokens": 133494598.0, "step": 3497 }, { "epoch": 0.44498155450960436, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.814902305603027, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8648212552070618, "num_tokens": 133536403.0, "step": 3498 }, { "epoch": 0.4451087647881949, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.793929100036621, "learning_rate": 1e-06, "loss": 0.5119, "mean_token_accuracy": 0.835155189037323, "num_tokens": 133567319.0, "step": 3499 }, { "epoch": 0.4452359750667854, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.79195499420166, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8493977189064026, "num_tokens": 133604829.0, "step": 3500 }, { "epoch": 0.4453631853453759, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.797027587890625, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8557485938072205, "num_tokens": 133638373.0, "step": 3501 }, { "epoch": 0.4454903956239664, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.785061836242676, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8667635917663574, "num_tokens": 133674110.0, "step": 3502 }, { "epoch": 0.44561760590255695, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.772085189819336, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8578680753707886, "num_tokens": 133711005.0, "step": 3503 }, { "epoch": 0.4457448161811474, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.764495849609375, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8690453767776489, "num_tokens": 133746258.0, "step": 3504 }, { "epoch": 0.44587202645973795, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.738097190856934, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8645936250686646, "num_tokens": 133790695.0, "step": 3505 }, { "epoch": 0.4459992367383285, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.763386726379395, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8529074788093567, "num_tokens": 133830181.0, "step": 3506 }, { "epoch": 0.44612644701691895, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.752490997314453, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8604075908660889, "num_tokens": 133866849.0, "step": 3507 }, { "epoch": 0.4462536572955095, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.803011894226074, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8560948371887207, "num_tokens": 133906812.0, "step": 3508 }, { "epoch": 0.4463808675741, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.787368774414062, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8494865894317627, "num_tokens": 133948956.0, "step": 3509 }, { "epoch": 0.4465080778526905, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.782247543334961, "learning_rate": 1e-06, "loss": 0.492, "mean_token_accuracy": 0.8393493890762329, "num_tokens": 133988764.0, "step": 3510 }, { "epoch": 0.446635288131281, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.796821594238281, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8608309626579285, "num_tokens": 134028148.0, "step": 3511 }, { "epoch": 0.44676249840987153, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.786306381225586, "learning_rate": 1e-06, "loss": 0.5233, "mean_token_accuracy": 0.8319193124771118, "num_tokens": 134065909.0, "step": 3512 }, { "epoch": 0.446889708688462, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.712014198303223, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8627328276634216, "num_tokens": 134108342.0, "step": 3513 }, { "epoch": 0.44701691896705253, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.736715316772461, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8552063703536987, "num_tokens": 134145928.0, "step": 3514 }, { "epoch": 0.44714412924564306, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.738926887512207, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8528254628181458, "num_tokens": 134185955.0, "step": 3515 }, { "epoch": 0.44727133952423354, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.717305183410645, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.861720621585846, "num_tokens": 134225829.0, "step": 3516 }, { "epoch": 0.44739854980282406, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.77536678314209, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8464404344558716, "num_tokens": 134261381.0, "step": 3517 }, { "epoch": 0.4475257600814146, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.759232521057129, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.863933801651001, "num_tokens": 134294697.0, "step": 3518 }, { "epoch": 0.44765297036000506, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.753325462341309, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8768174648284912, "num_tokens": 134334349.0, "step": 3519 }, { "epoch": 0.4477801806385956, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.774923324584961, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8450880646705627, "num_tokens": 134373381.0, "step": 3520 }, { "epoch": 0.4479073909171861, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.757416725158691, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.861577033996582, "num_tokens": 134407952.0, "step": 3521 }, { "epoch": 0.4480346011957766, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.759204864501953, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8468747138977051, "num_tokens": 134458862.0, "step": 3522 }, { "epoch": 0.4481618114743671, "ewc_loss": 0.01611328125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 9.74122142791748, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8528011441230774, "num_tokens": 134494007.0, "step": 3523 }, { "epoch": 0.44828902175295765, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.76925277709961, "learning_rate": 1e-06, "loss": 0.4889, "mean_token_accuracy": 0.8419573307037354, "num_tokens": 134534195.0, "step": 3524 }, { "epoch": 0.4484162320315481, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.76096248626709, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8642999529838562, "num_tokens": 134570241.0, "step": 3525 }, { "epoch": 0.44854344231013865, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.739014625549316, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8544237017631531, "num_tokens": 134611904.0, "step": 3526 }, { "epoch": 0.4486706525887292, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.823079109191895, "learning_rate": 1e-06, "loss": 0.4823, "mean_token_accuracy": 0.8466068506240845, "num_tokens": 134644426.0, "step": 3527 }, { "epoch": 0.44879786286731965, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.795110702514648, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8643511533737183, "num_tokens": 134678979.0, "step": 3528 }, { "epoch": 0.4489250731459102, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.792574882507324, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8586204648017883, "num_tokens": 134717270.0, "step": 3529 }, { "epoch": 0.4490522834245007, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.762797355651855, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.856168270111084, "num_tokens": 134755464.0, "step": 3530 }, { "epoch": 0.44917949370309124, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.786571502685547, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8479207754135132, "num_tokens": 134793737.0, "step": 3531 }, { "epoch": 0.4493067039816817, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.76717758178711, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8511018753051758, "num_tokens": 134834332.0, "step": 3532 }, { "epoch": 0.44943391426027224, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.785003662109375, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8665170669555664, "num_tokens": 134871027.0, "step": 3533 }, { "epoch": 0.44956112453886277, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.77971363067627, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8567080497741699, "num_tokens": 134908529.0, "step": 3534 }, { "epoch": 0.44968833481745324, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.785767555236816, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8650869131088257, "num_tokens": 134944177.0, "step": 3535 }, { "epoch": 0.44981554509604377, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.757428169250488, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8548581600189209, "num_tokens": 134977928.0, "step": 3536 }, { "epoch": 0.4499427553746343, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.808737754821777, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8568308353424072, "num_tokens": 135017554.0, "step": 3537 }, { "epoch": 0.45006996565322477, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.748268127441406, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8627227544784546, "num_tokens": 135058694.0, "step": 3538 }, { "epoch": 0.4501971759318153, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.803107261657715, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8582766652107239, "num_tokens": 135094037.0, "step": 3539 }, { "epoch": 0.4503243862104058, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.831727027893066, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8503586649894714, "num_tokens": 135132758.0, "step": 3540 }, { "epoch": 0.4504515964889963, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.786696434020996, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.868749737739563, "num_tokens": 135169398.0, "step": 3541 }, { "epoch": 0.4505788067675868, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.827157974243164, "learning_rate": 1e-06, "loss": 0.5147, "mean_token_accuracy": 0.842835545539856, "num_tokens": 135204312.0, "step": 3542 }, { "epoch": 0.45070601704617735, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.77181625366211, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8540017008781433, "num_tokens": 135246287.0, "step": 3543 }, { "epoch": 0.4508332273247678, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.775594711303711, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8595799803733826, "num_tokens": 135280970.0, "step": 3544 }, { "epoch": 0.45096043760335836, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.755086898803711, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8787693977355957, "num_tokens": 135317873.0, "step": 3545 }, { "epoch": 0.4510876478819489, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.754470825195312, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8631558418273926, "num_tokens": 135354994.0, "step": 3546 }, { "epoch": 0.45121485816053936, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.78299331665039, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8513860106468201, "num_tokens": 135393050.0, "step": 3547 }, { "epoch": 0.4513420684391299, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.78514289855957, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8614304065704346, "num_tokens": 135431704.0, "step": 3548 }, { "epoch": 0.4514692787177204, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.811777114868164, "learning_rate": 1e-06, "loss": 0.5082, "mean_token_accuracy": 0.8341062068939209, "num_tokens": 135471306.0, "step": 3549 }, { "epoch": 0.4515964889963109, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.804747581481934, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8435741662979126, "num_tokens": 135509647.0, "step": 3550 }, { "epoch": 0.4517236992749014, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.802433013916016, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8643255233764648, "num_tokens": 135550469.0, "step": 3551 }, { "epoch": 0.45185090955349194, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.793046951293945, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8566960096359253, "num_tokens": 135589290.0, "step": 3552 }, { "epoch": 0.4519781198320824, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.799034118652344, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8514149188995361, "num_tokens": 135624416.0, "step": 3553 }, { "epoch": 0.45210533011067294, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.774493217468262, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8593093156814575, "num_tokens": 135662316.0, "step": 3554 }, { "epoch": 0.45223254038926347, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.827290534973145, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8578999638557434, "num_tokens": 135701739.0, "step": 3555 }, { "epoch": 0.45235975066785394, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.802431106567383, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8524351119995117, "num_tokens": 135741372.0, "step": 3556 }, { "epoch": 0.45248696094644447, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.766115188598633, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8779164552688599, "num_tokens": 135779945.0, "step": 3557 }, { "epoch": 0.452614171225035, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.828882217407227, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8508967161178589, "num_tokens": 135813179.0, "step": 3558 }, { "epoch": 0.4527413815036255, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.82280445098877, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8590331077575684, "num_tokens": 135855147.0, "step": 3559 }, { "epoch": 0.452868591782216, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.778302192687988, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8642278909683228, "num_tokens": 135892990.0, "step": 3560 }, { "epoch": 0.45299580206080653, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.792614936828613, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8749220371246338, "num_tokens": 135926911.0, "step": 3561 }, { "epoch": 0.453123012339397, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.839898109436035, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8651260137557983, "num_tokens": 135963530.0, "step": 3562 }, { "epoch": 0.45325022261798753, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.806280136108398, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8475605249404907, "num_tokens": 136000677.0, "step": 3563 }, { "epoch": 0.45337743289657806, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.835439682006836, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8456476926803589, "num_tokens": 136037283.0, "step": 3564 }, { "epoch": 0.45350464317516853, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.855602264404297, "learning_rate": 1e-06, "loss": 0.501, "mean_token_accuracy": 0.8394742012023926, "num_tokens": 136071454.0, "step": 3565 }, { "epoch": 0.45363185345375906, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.817524909973145, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8458069562911987, "num_tokens": 136105638.0, "step": 3566 }, { "epoch": 0.4537590637323496, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.741449356079102, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8602118492126465, "num_tokens": 136147968.0, "step": 3567 }, { "epoch": 0.45388627401094006, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.879281044006348, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8569651246070862, "num_tokens": 136190468.0, "step": 3568 }, { "epoch": 0.4540134842895306, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.820320129394531, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8607882857322693, "num_tokens": 136229363.0, "step": 3569 }, { "epoch": 0.4541406945681211, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.86719036102295, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8503833413124084, "num_tokens": 136268834.0, "step": 3570 }, { "epoch": 0.4542679048467116, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.813218116760254, "learning_rate": 1e-06, "loss": 0.5278, "mean_token_accuracy": 0.8337053060531616, "num_tokens": 136306761.0, "step": 3571 }, { "epoch": 0.4543951151253021, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.793405532836914, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8558983206748962, "num_tokens": 136347736.0, "step": 3572 }, { "epoch": 0.45452232540389265, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.846253395080566, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8599861860275269, "num_tokens": 136381009.0, "step": 3573 }, { "epoch": 0.4546495356824831, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 9.736754417419434, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8672868609428406, "num_tokens": 136418050.0, "step": 3574 }, { "epoch": 0.45477674596107365, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.834043502807617, "learning_rate": 1e-06, "loss": 0.5014, "mean_token_accuracy": 0.838010847568512, "num_tokens": 136459109.0, "step": 3575 }, { "epoch": 0.4549039562396642, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.780054092407227, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8524491786956787, "num_tokens": 136497993.0, "step": 3576 }, { "epoch": 0.45503116651825465, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.804036140441895, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8589503765106201, "num_tokens": 136534446.0, "step": 3577 }, { "epoch": 0.4551583767968452, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.835953712463379, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8408260345458984, "num_tokens": 136571117.0, "step": 3578 }, { "epoch": 0.4552855870754357, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.815412521362305, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8640434741973877, "num_tokens": 136601706.0, "step": 3579 }, { "epoch": 0.4554127973540262, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.798458099365234, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8566477298736572, "num_tokens": 136639735.0, "step": 3580 }, { "epoch": 0.4555400076326167, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.79013442993164, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8495007753372192, "num_tokens": 136675647.0, "step": 3581 }, { "epoch": 0.45566721791120723, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.815879821777344, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8645225167274475, "num_tokens": 136713361.0, "step": 3582 }, { "epoch": 0.45579442818979776, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.764025688171387, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8556034564971924, "num_tokens": 136751440.0, "step": 3583 }, { "epoch": 0.45592163846838824, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.808162689208984, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8522864580154419, "num_tokens": 136786078.0, "step": 3584 }, { "epoch": 0.45604884874697876, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.823479652404785, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8640580773353577, "num_tokens": 136828313.0, "step": 3585 }, { "epoch": 0.4561760590255693, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.835660934448242, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8554388284683228, "num_tokens": 136862549.0, "step": 3586 }, { "epoch": 0.45630326930415976, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.792488098144531, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8592209815979004, "num_tokens": 136906423.0, "step": 3587 }, { "epoch": 0.4564304795827503, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.819127082824707, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8453851938247681, "num_tokens": 136941550.0, "step": 3588 }, { "epoch": 0.4565576898613408, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.820317268371582, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8595903515815735, "num_tokens": 136980234.0, "step": 3589 }, { "epoch": 0.4566849001399313, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.785707473754883, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8603018522262573, "num_tokens": 137023991.0, "step": 3590 }, { "epoch": 0.4568121104185218, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.826371192932129, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8476942777633667, "num_tokens": 137063261.0, "step": 3591 }, { "epoch": 0.45693932069711235, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.831551551818848, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8650253415107727, "num_tokens": 137103935.0, "step": 3592 }, { "epoch": 0.4570665309757028, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.778881072998047, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8568105101585388, "num_tokens": 137140194.0, "step": 3593 }, { "epoch": 0.45719374125429335, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.840117454528809, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.8476496338844299, "num_tokens": 137179086.0, "step": 3594 }, { "epoch": 0.4573209515328839, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.80872631072998, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8623881340026855, "num_tokens": 137216887.0, "step": 3595 }, { "epoch": 0.45744816181147435, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.827679634094238, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8724198937416077, "num_tokens": 137255029.0, "step": 3596 }, { "epoch": 0.4575753720900649, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.807572364807129, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.860605001449585, "num_tokens": 137291399.0, "step": 3597 }, { "epoch": 0.4577025823686554, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.85006046295166, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8612419366836548, "num_tokens": 137326569.0, "step": 3598 }, { "epoch": 0.4578297926472459, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.787198066711426, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8527826070785522, "num_tokens": 137370035.0, "step": 3599 }, { "epoch": 0.4579570029258364, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.815142631530762, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8528395891189575, "num_tokens": 137407969.0, "step": 3600 }, { "epoch": 0.45808421320442694, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.7923002243042, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8552329540252686, "num_tokens": 137447256.0, "step": 3601 }, { "epoch": 0.4582114234830174, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.897637367248535, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.8419753313064575, "num_tokens": 137479762.0, "step": 3602 }, { "epoch": 0.45833863376160794, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.803829193115234, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8495380878448486, "num_tokens": 137521823.0, "step": 3603 }, { "epoch": 0.45846584404019847, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.795907020568848, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8538918495178223, "num_tokens": 137558174.0, "step": 3604 }, { "epoch": 0.45859305431878894, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.835256576538086, "learning_rate": 1e-06, "loss": 0.5282, "mean_token_accuracy": 0.8324555158615112, "num_tokens": 137597265.0, "step": 3605 }, { "epoch": 0.45872026459737947, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.768345832824707, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8639887571334839, "num_tokens": 137635954.0, "step": 3606 }, { "epoch": 0.45884747487597, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.827704429626465, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8589604496955872, "num_tokens": 137676137.0, "step": 3607 }, { "epoch": 0.45897468515456047, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.775897979736328, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8593335151672363, "num_tokens": 137716723.0, "step": 3608 }, { "epoch": 0.459101895433151, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.824416160583496, "learning_rate": 1e-06, "loss": 0.4955, "mean_token_accuracy": 0.8390582799911499, "num_tokens": 137752435.0, "step": 3609 }, { "epoch": 0.4592291057117415, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.753111839294434, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8478001356124878, "num_tokens": 137786684.0, "step": 3610 }, { "epoch": 0.459356315990332, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.810089111328125, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8479353785514832, "num_tokens": 137830668.0, "step": 3611 }, { "epoch": 0.4594835262689225, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.770071983337402, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8625490665435791, "num_tokens": 137865262.0, "step": 3612 }, { "epoch": 0.45961073654751305, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.827232360839844, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.863173246383667, "num_tokens": 137898867.0, "step": 3613 }, { "epoch": 0.4597379468261035, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.810036659240723, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8579013347625732, "num_tokens": 137935108.0, "step": 3614 }, { "epoch": 0.45986515710469406, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.80047607421875, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8674395084381104, "num_tokens": 137977714.0, "step": 3615 }, { "epoch": 0.4599923673832846, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.838485717773438, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.868363618850708, "num_tokens": 138006345.0, "step": 3616 }, { "epoch": 0.46011957766187506, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.80595874786377, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8570796847343445, "num_tokens": 138045106.0, "step": 3617 }, { "epoch": 0.4602467879404656, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.81579303741455, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8511209487915039, "num_tokens": 138079993.0, "step": 3618 }, { "epoch": 0.4603739982190561, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.799778938293457, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8605881929397583, "num_tokens": 138118555.0, "step": 3619 }, { "epoch": 0.4605012084976466, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.80876636505127, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8649085760116577, "num_tokens": 138155096.0, "step": 3620 }, { "epoch": 0.4606284187762371, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.764073371887207, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8589360117912292, "num_tokens": 138194529.0, "step": 3621 }, { "epoch": 0.46075562905482764, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.845662117004395, "learning_rate": 1e-06, "loss": 0.5164, "mean_token_accuracy": 0.8316928744316101, "num_tokens": 138235520.0, "step": 3622 }, { "epoch": 0.4608828393334181, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.79665756225586, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8670374155044556, "num_tokens": 138271391.0, "step": 3623 }, { "epoch": 0.46101004961200864, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.818612098693848, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8510134220123291, "num_tokens": 138313923.0, "step": 3624 }, { "epoch": 0.46113725989059917, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.816969871520996, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.8489737510681152, "num_tokens": 138355992.0, "step": 3625 }, { "epoch": 0.46126447016918964, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.874938011169434, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8774193525314331, "num_tokens": 138393074.0, "step": 3626 }, { "epoch": 0.4613916804477802, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.7982759475708, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8633415102958679, "num_tokens": 138427854.0, "step": 3627 }, { "epoch": 0.4615188907263707, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.840954780578613, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8667657375335693, "num_tokens": 138467359.0, "step": 3628 }, { "epoch": 0.4616461010049612, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.802528381347656, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8456737995147705, "num_tokens": 138505355.0, "step": 3629 }, { "epoch": 0.4617733112835517, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.821674346923828, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8611571788787842, "num_tokens": 138543703.0, "step": 3630 }, { "epoch": 0.46190052156214223, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.808431625366211, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8497241735458374, "num_tokens": 138579860.0, "step": 3631 }, { "epoch": 0.46202773184073276, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.85319995880127, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8664226531982422, "num_tokens": 138615755.0, "step": 3632 }, { "epoch": 0.46215494211932323, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.832921981811523, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8525573015213013, "num_tokens": 138647607.0, "step": 3633 }, { "epoch": 0.46228215239791376, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.842768669128418, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8623514175415039, "num_tokens": 138679284.0, "step": 3634 }, { "epoch": 0.4624093626765043, "ewc_loss": 0.016357421875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.787330627441406, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8665663003921509, "num_tokens": 138719690.0, "step": 3635 }, { "epoch": 0.46253657295509476, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.829361915588379, "learning_rate": 1e-06, "loss": 0.4825, "mean_token_accuracy": 0.8455622792243958, "num_tokens": 138758031.0, "step": 3636 }, { "epoch": 0.4626637832336853, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.84699821472168, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8569657206535339, "num_tokens": 138797283.0, "step": 3637 }, { "epoch": 0.4627909935122758, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.796451568603516, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8533070087432861, "num_tokens": 138834672.0, "step": 3638 }, { "epoch": 0.4629182037908663, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.834561347961426, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8679566383361816, "num_tokens": 138871105.0, "step": 3639 }, { "epoch": 0.4630454140694568, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.829105377197266, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8583942651748657, "num_tokens": 138903516.0, "step": 3640 }, { "epoch": 0.46317262434804735, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.846301078796387, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8555213212966919, "num_tokens": 138942021.0, "step": 3641 }, { "epoch": 0.4632998346266378, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.80170726776123, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8658912181854248, "num_tokens": 138978210.0, "step": 3642 }, { "epoch": 0.46342704490522835, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.828402519226074, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8599947690963745, "num_tokens": 139019122.0, "step": 3643 }, { "epoch": 0.4635542551838189, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.829407691955566, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8675109148025513, "num_tokens": 139062187.0, "step": 3644 }, { "epoch": 0.46368146546240935, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.805612564086914, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8756096363067627, "num_tokens": 139102960.0, "step": 3645 }, { "epoch": 0.4638086757409999, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.85263729095459, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8585929870605469, "num_tokens": 139137302.0, "step": 3646 }, { "epoch": 0.4639358860195904, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.83376693725586, "learning_rate": 1e-06, "loss": 0.508, "mean_token_accuracy": 0.8378530144691467, "num_tokens": 139177195.0, "step": 3647 }, { "epoch": 0.4640630962981809, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.826295852661133, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8540263175964355, "num_tokens": 139214722.0, "step": 3648 }, { "epoch": 0.4641903065767714, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.852375984191895, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8556627035140991, "num_tokens": 139248569.0, "step": 3649 }, { "epoch": 0.46431751685536193, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.854523658752441, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8638036251068115, "num_tokens": 139289141.0, "step": 3650 }, { "epoch": 0.4644447271339524, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.782452583312988, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8748482465744019, "num_tokens": 139330791.0, "step": 3651 }, { "epoch": 0.46457193741254293, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.887866973876953, "learning_rate": 1e-06, "loss": 0.5392, "mean_token_accuracy": 0.8258789777755737, "num_tokens": 139373009.0, "step": 3652 }, { "epoch": 0.46469914769113346, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.815223693847656, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8562244176864624, "num_tokens": 139408894.0, "step": 3653 }, { "epoch": 0.46482635796972394, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.845134735107422, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8558706641197205, "num_tokens": 139444040.0, "step": 3654 }, { "epoch": 0.46495356824831446, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.807206153869629, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8629999756813049, "num_tokens": 139479675.0, "step": 3655 }, { "epoch": 0.465080778526905, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.811503410339355, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8647993803024292, "num_tokens": 139521792.0, "step": 3656 }, { "epoch": 0.46520798880549546, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.813138008117676, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8646347522735596, "num_tokens": 139563471.0, "step": 3657 }, { "epoch": 0.465335199084086, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.858057022094727, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8582720756530762, "num_tokens": 139604702.0, "step": 3658 }, { "epoch": 0.4654624093626765, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.851902961730957, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8554985523223877, "num_tokens": 139644573.0, "step": 3659 }, { "epoch": 0.465589619641267, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.779620170593262, "learning_rate": 1e-06, "loss": 0.4846, "mean_token_accuracy": 0.842860996723175, "num_tokens": 139680924.0, "step": 3660 }, { "epoch": 0.4657168299198575, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.846871376037598, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.849913477897644, "num_tokens": 139724369.0, "step": 3661 }, { "epoch": 0.46584404019844805, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.784893035888672, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8581721186637878, "num_tokens": 139769412.0, "step": 3662 }, { "epoch": 0.4659712504770385, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.897400856018066, "learning_rate": 1e-06, "loss": 0.5133, "mean_token_accuracy": 0.8333131074905396, "num_tokens": 139808066.0, "step": 3663 }, { "epoch": 0.46609846075562905, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.81400203704834, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8632025718688965, "num_tokens": 139851169.0, "step": 3664 }, { "epoch": 0.4662256710342196, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.855177879333496, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8674396276473999, "num_tokens": 139890550.0, "step": 3665 }, { "epoch": 0.46635288131281005, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.818578720092773, "learning_rate": 1e-06, "loss": 0.5225, "mean_token_accuracy": 0.8311378955841064, "num_tokens": 139931496.0, "step": 3666 }, { "epoch": 0.4664800915914006, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 9.838007926940918, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.8481535315513611, "num_tokens": 139965872.0, "step": 3667 }, { "epoch": 0.4666073018699911, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.889328002929688, "learning_rate": 1e-06, "loss": 0.512, "mean_token_accuracy": 0.842272162437439, "num_tokens": 140000394.0, "step": 3668 }, { "epoch": 0.4667345121485816, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.785201072692871, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.8405687212944031, "num_tokens": 140039967.0, "step": 3669 }, { "epoch": 0.4668617224271721, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.820683479309082, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8575732707977295, "num_tokens": 140075739.0, "step": 3670 }, { "epoch": 0.46698893270576264, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.835352897644043, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8622589707374573, "num_tokens": 140110016.0, "step": 3671 }, { "epoch": 0.4671161429843531, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 9.856801986694336, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8470357656478882, "num_tokens": 140147905.0, "step": 3672 }, { "epoch": 0.46724335326294364, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.862046241760254, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8518039584159851, "num_tokens": 140182284.0, "step": 3673 }, { "epoch": 0.46737056354153417, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.839010238647461, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8497431874275208, "num_tokens": 140220669.0, "step": 3674 }, { "epoch": 0.46749777382012464, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.823650360107422, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8430098295211792, "num_tokens": 140258399.0, "step": 3675 }, { "epoch": 0.46762498409871517, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.836628913879395, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8709449768066406, "num_tokens": 140296015.0, "step": 3676 }, { "epoch": 0.4677521943773057, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.818148612976074, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8462170362472534, "num_tokens": 140338618.0, "step": 3677 }, { "epoch": 0.46787940465589617, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.822318077087402, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8611648082733154, "num_tokens": 140379151.0, "step": 3678 }, { "epoch": 0.4680066149344867, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.806655883789062, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8430586457252502, "num_tokens": 140417740.0, "step": 3679 }, { "epoch": 0.4681338252130772, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.839503288269043, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8544202446937561, "num_tokens": 140452454.0, "step": 3680 }, { "epoch": 0.46826103549166775, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.863348007202148, "learning_rate": 1e-06, "loss": 0.5221, "mean_token_accuracy": 0.8343102931976318, "num_tokens": 140492981.0, "step": 3681 }, { "epoch": 0.4683882457702582, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.824969291687012, "learning_rate": 1e-06, "loss": 0.4747, "mean_token_accuracy": 0.8460694551467896, "num_tokens": 140534131.0, "step": 3682 }, { "epoch": 0.46851545604884876, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.866498947143555, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8495182991027832, "num_tokens": 140576761.0, "step": 3683 }, { "epoch": 0.4686426663274393, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.859123229980469, "learning_rate": 1e-06, "loss": 0.5308, "mean_token_accuracy": 0.8370569944381714, "num_tokens": 140609538.0, "step": 3684 }, { "epoch": 0.46876987660602976, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.91198444366455, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8516459465026855, "num_tokens": 140641802.0, "step": 3685 }, { "epoch": 0.4688970868846203, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.788986206054688, "learning_rate": 1e-06, "loss": 0.4865, "mean_token_accuracy": 0.8448019027709961, "num_tokens": 140685832.0, "step": 3686 }, { "epoch": 0.4690242971632108, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.895977973937988, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8442715406417847, "num_tokens": 140725333.0, "step": 3687 }, { "epoch": 0.4691515074418013, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.906472206115723, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8533031344413757, "num_tokens": 140760838.0, "step": 3688 }, { "epoch": 0.4692787177203918, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.84123420715332, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8453488349914551, "num_tokens": 140797367.0, "step": 3689 }, { "epoch": 0.46940592799898234, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.861207962036133, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8538342118263245, "num_tokens": 140839778.0, "step": 3690 }, { "epoch": 0.4695331382775728, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.826545715332031, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8512744903564453, "num_tokens": 140880340.0, "step": 3691 }, { "epoch": 0.46966034855616334, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.847543716430664, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.848914384841919, "num_tokens": 140915451.0, "step": 3692 }, { "epoch": 0.46978755883475387, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.825953483581543, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8527910709381104, "num_tokens": 140956967.0, "step": 3693 }, { "epoch": 0.46991476911334434, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.869362831115723, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8701202869415283, "num_tokens": 140996640.0, "step": 3694 }, { "epoch": 0.47004197939193487, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.8070068359375, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8543616533279419, "num_tokens": 141036094.0, "step": 3695 }, { "epoch": 0.4701691896705254, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.895951271057129, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8448717594146729, "num_tokens": 141069033.0, "step": 3696 }, { "epoch": 0.4702963999491159, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.82144546508789, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.872568666934967, "num_tokens": 141111075.0, "step": 3697 }, { "epoch": 0.4704236102277064, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.826180458068848, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8709406852722168, "num_tokens": 141145270.0, "step": 3698 }, { "epoch": 0.47055082050629693, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.872502326965332, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8702551126480103, "num_tokens": 141182132.0, "step": 3699 }, { "epoch": 0.4706780307848874, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.875840187072754, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.856067955493927, "num_tokens": 141217455.0, "step": 3700 }, { "epoch": 0.47080524106347793, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.865421295166016, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8547496795654297, "num_tokens": 141249680.0, "step": 3701 }, { "epoch": 0.47093245134206846, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.840376853942871, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8554732799530029, "num_tokens": 141292085.0, "step": 3702 }, { "epoch": 0.47105966162065893, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.816374778747559, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8642265200614929, "num_tokens": 141331455.0, "step": 3703 }, { "epoch": 0.47118687189924946, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.859651565551758, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8555211424827576, "num_tokens": 141365623.0, "step": 3704 }, { "epoch": 0.47131408217784, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.855817794799805, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8476731777191162, "num_tokens": 141399033.0, "step": 3705 }, { "epoch": 0.47144129245643046, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.862157821655273, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8632611036300659, "num_tokens": 141430436.0, "step": 3706 }, { "epoch": 0.471568502735021, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.861532211303711, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.872843861579895, "num_tokens": 141463169.0, "step": 3707 }, { "epoch": 0.4716957130136115, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.836978912353516, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8747860789299011, "num_tokens": 141495524.0, "step": 3708 }, { "epoch": 0.471822923292202, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.92306137084961, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8679354190826416, "num_tokens": 141533099.0, "step": 3709 }, { "epoch": 0.4719501335707925, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.774016380310059, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.86127769947052, "num_tokens": 141576389.0, "step": 3710 }, { "epoch": 0.47207734384938305, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.877740859985352, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8589813113212585, "num_tokens": 141611870.0, "step": 3711 }, { "epoch": 0.4722045541279735, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.941082000732422, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8680834770202637, "num_tokens": 141648539.0, "step": 3712 }, { "epoch": 0.47233176440656405, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.815306663513184, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8603341579437256, "num_tokens": 141688549.0, "step": 3713 }, { "epoch": 0.4724589746851546, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.855408668518066, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8658723831176758, "num_tokens": 141725294.0, "step": 3714 }, { "epoch": 0.47258618496374505, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.873031616210938, "learning_rate": 1e-06, "loss": 0.4888, "mean_token_accuracy": 0.8413037061691284, "num_tokens": 141756761.0, "step": 3715 }, { "epoch": 0.4727133952423356, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.876258850097656, "learning_rate": 1e-06, "loss": 0.4942, "mean_token_accuracy": 0.8420237302780151, "num_tokens": 141790417.0, "step": 3716 }, { "epoch": 0.4728406055209261, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.833442687988281, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8679153323173523, "num_tokens": 141830749.0, "step": 3717 }, { "epoch": 0.4729678157995166, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.884461402893066, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8610014915466309, "num_tokens": 141868433.0, "step": 3718 }, { "epoch": 0.4730950260781071, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.795783042907715, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8751516342163086, "num_tokens": 141907140.0, "step": 3719 }, { "epoch": 0.47322223635669763, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.82898998260498, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8710427284240723, "num_tokens": 141947436.0, "step": 3720 }, { "epoch": 0.4733494466352881, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.840230941772461, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8510648012161255, "num_tokens": 141986876.0, "step": 3721 }, { "epoch": 0.47347665691387864, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.850053787231445, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8614596724510193, "num_tokens": 142025806.0, "step": 3722 }, { "epoch": 0.47360386719246916, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.86898422241211, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8561468124389648, "num_tokens": 142057932.0, "step": 3723 }, { "epoch": 0.47373107747105964, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.869645118713379, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8602452278137207, "num_tokens": 142088880.0, "step": 3724 }, { "epoch": 0.47385828774965016, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.913337707519531, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.870807409286499, "num_tokens": 142124959.0, "step": 3725 }, { "epoch": 0.4739854980282407, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.942132949829102, "learning_rate": 1e-06, "loss": 0.5055, "mean_token_accuracy": 0.8430984020233154, "num_tokens": 142160667.0, "step": 3726 }, { "epoch": 0.47411270830683117, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.872225761413574, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8568891286849976, "num_tokens": 142205017.0, "step": 3727 }, { "epoch": 0.4742399185854217, "ewc_loss": 0.0166015625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 9.79610824584961, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8639235496520996, "num_tokens": 142241479.0, "step": 3728 }, { "epoch": 0.4743671288640122, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.846123695373535, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8582085371017456, "num_tokens": 142280881.0, "step": 3729 }, { "epoch": 0.4744943391426027, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.821094512939453, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8655558228492737, "num_tokens": 142328248.0, "step": 3730 }, { "epoch": 0.4746215494211932, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.90648365020752, "learning_rate": 1e-06, "loss": 0.5061, "mean_token_accuracy": 0.8416075110435486, "num_tokens": 142364872.0, "step": 3731 }, { "epoch": 0.47474875969978375, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.91240119934082, "learning_rate": 1e-06, "loss": 0.4979, "mean_token_accuracy": 0.8397263288497925, "num_tokens": 142400494.0, "step": 3732 }, { "epoch": 0.4748759699783743, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.894022941589355, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8547786474227905, "num_tokens": 142433739.0, "step": 3733 }, { "epoch": 0.47500318025696475, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.904088020324707, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8660883903503418, "num_tokens": 142469777.0, "step": 3734 }, { "epoch": 0.4751303905355553, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.821724891662598, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8471885323524475, "num_tokens": 142510052.0, "step": 3735 }, { "epoch": 0.4752576008141458, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.888914108276367, "learning_rate": 1e-06, "loss": 0.4853, "mean_token_accuracy": 0.8453973531723022, "num_tokens": 142545689.0, "step": 3736 }, { "epoch": 0.4753848110927363, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.840459823608398, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8568793535232544, "num_tokens": 142585916.0, "step": 3737 }, { "epoch": 0.4755120213713268, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.857048988342285, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8644264936447144, "num_tokens": 142623019.0, "step": 3738 }, { "epoch": 0.47563923164991734, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.860980987548828, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8538024425506592, "num_tokens": 142660986.0, "step": 3739 }, { "epoch": 0.4757664419285078, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.807361602783203, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.860687255859375, "num_tokens": 142702876.0, "step": 3740 }, { "epoch": 0.47589365220709834, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.858558654785156, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8461617231369019, "num_tokens": 142744082.0, "step": 3741 }, { "epoch": 0.47602086248568887, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.873824119567871, "learning_rate": 1e-06, "loss": 0.4943, "mean_token_accuracy": 0.8389859199523926, "num_tokens": 142777511.0, "step": 3742 }, { "epoch": 0.47614807276427934, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.890766143798828, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8520689010620117, "num_tokens": 142818882.0, "step": 3743 }, { "epoch": 0.47627528304286987, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.921076774597168, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8600156307220459, "num_tokens": 142852771.0, "step": 3744 }, { "epoch": 0.4764024933214604, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.85647964477539, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8578143119812012, "num_tokens": 142888398.0, "step": 3745 }, { "epoch": 0.47652970360005087, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.947673797607422, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8726053833961487, "num_tokens": 142924789.0, "step": 3746 }, { "epoch": 0.4766569138786414, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.883255004882812, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8649687767028809, "num_tokens": 142958843.0, "step": 3747 }, { "epoch": 0.4767841241572319, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.846575736999512, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8678333759307861, "num_tokens": 142996408.0, "step": 3748 }, { "epoch": 0.4769113344358224, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.88621997833252, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8501745462417603, "num_tokens": 143040031.0, "step": 3749 }, { "epoch": 0.4770385447144129, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.835832595825195, "learning_rate": 1e-06, "loss": 0.5097, "mean_token_accuracy": 0.8348662257194519, "num_tokens": 143081648.0, "step": 3750 }, { "epoch": 0.47716575499300345, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.865684509277344, "learning_rate": 1e-06, "loss": 0.5037, "mean_token_accuracy": 0.8367899060249329, "num_tokens": 143118913.0, "step": 3751 }, { "epoch": 0.4772929652715939, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.854601860046387, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8664341568946838, "num_tokens": 143153858.0, "step": 3752 }, { "epoch": 0.47742017555018446, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.8970365524292, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8681984543800354, "num_tokens": 143190977.0, "step": 3753 }, { "epoch": 0.477547385828775, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.869851112365723, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8495434522628784, "num_tokens": 143228009.0, "step": 3754 }, { "epoch": 0.47767459610736546, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.851716995239258, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8465216159820557, "num_tokens": 143269820.0, "step": 3755 }, { "epoch": 0.477801806385956, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.843073844909668, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8608589172363281, "num_tokens": 143303441.0, "step": 3756 }, { "epoch": 0.4779290166645465, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.879213333129883, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8487982749938965, "num_tokens": 143338748.0, "step": 3757 }, { "epoch": 0.478056226943137, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.830785751342773, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.857680082321167, "num_tokens": 143373900.0, "step": 3758 }, { "epoch": 0.4781834372217275, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.891050338745117, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8523749113082886, "num_tokens": 143411190.0, "step": 3759 }, { "epoch": 0.47831064750031804, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.846977233886719, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.852501630783081, "num_tokens": 143449470.0, "step": 3760 }, { "epoch": 0.4784378577789085, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.83167839050293, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8634382486343384, "num_tokens": 143485868.0, "step": 3761 }, { "epoch": 0.47856506805749904, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.869202613830566, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8660526275634766, "num_tokens": 143520334.0, "step": 3762 }, { "epoch": 0.47869227833608957, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.884208679199219, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.850693941116333, "num_tokens": 143557752.0, "step": 3763 }, { "epoch": 0.47881948861468004, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.895977973937988, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8565425872802734, "num_tokens": 143593385.0, "step": 3764 }, { "epoch": 0.4789466988932706, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.85891056060791, "learning_rate": 1e-06, "loss": 0.5092, "mean_token_accuracy": 0.837142825126648, "num_tokens": 143631975.0, "step": 3765 }, { "epoch": 0.4790739091718611, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.926286697387695, "learning_rate": 1e-06, "loss": 0.4931, "mean_token_accuracy": 0.8393141627311707, "num_tokens": 143668382.0, "step": 3766 }, { "epoch": 0.4792011194504516, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.873480796813965, "learning_rate": 1e-06, "loss": 0.486, "mean_token_accuracy": 0.8422569036483765, "num_tokens": 143706623.0, "step": 3767 }, { "epoch": 0.4793283297290421, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.890764236450195, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8611820936203003, "num_tokens": 143745664.0, "step": 3768 }, { "epoch": 0.47945554000763263, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.85470962524414, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8473392724990845, "num_tokens": 143780761.0, "step": 3769 }, { "epoch": 0.4795827502862231, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.882509231567383, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8613977432250977, "num_tokens": 143825298.0, "step": 3770 }, { "epoch": 0.47970996056481363, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.875666618347168, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8765811920166016, "num_tokens": 143859855.0, "step": 3771 }, { "epoch": 0.47983717084340416, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.878694534301758, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8700553774833679, "num_tokens": 143898553.0, "step": 3772 }, { "epoch": 0.47996438112199463, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.942051887512207, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8543410301208496, "num_tokens": 143935362.0, "step": 3773 }, { "epoch": 0.48009159140058516, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.872759819030762, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8576145172119141, "num_tokens": 143978692.0, "step": 3774 }, { "epoch": 0.4802188016791757, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.892914772033691, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8646388053894043, "num_tokens": 144020951.0, "step": 3775 }, { "epoch": 0.48034601195776616, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.887056350708008, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8658764958381653, "num_tokens": 144058341.0, "step": 3776 }, { "epoch": 0.4804732222363567, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.893181800842285, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8494296073913574, "num_tokens": 144098232.0, "step": 3777 }, { "epoch": 0.4806004325149472, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.907448768615723, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8517276644706726, "num_tokens": 144137537.0, "step": 3778 }, { "epoch": 0.4807276427935377, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.87491226196289, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8618484735488892, "num_tokens": 144178459.0, "step": 3779 }, { "epoch": 0.4808548530721282, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.910429000854492, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8513350486755371, "num_tokens": 144219257.0, "step": 3780 }, { "epoch": 0.48098206335071875, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.869877815246582, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8632142543792725, "num_tokens": 144262232.0, "step": 3781 }, { "epoch": 0.4811092736293093, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.908828735351562, "learning_rate": 1e-06, "loss": 0.5018, "mean_token_accuracy": 0.8351365327835083, "num_tokens": 144297144.0, "step": 3782 }, { "epoch": 0.48123648390789975, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.933304786682129, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8578065633773804, "num_tokens": 144336078.0, "step": 3783 }, { "epoch": 0.4813636941864903, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.886231422424316, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8494791984558105, "num_tokens": 144380162.0, "step": 3784 }, { "epoch": 0.4814909044650808, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.914188385009766, "learning_rate": 1e-06, "loss": 0.4848, "mean_token_accuracy": 0.8428477048873901, "num_tokens": 144416735.0, "step": 3785 }, { "epoch": 0.4816181147436713, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.94205093383789, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8463865518569946, "num_tokens": 144452780.0, "step": 3786 }, { "epoch": 0.4817453250222618, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.843514442443848, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8625994920730591, "num_tokens": 144494155.0, "step": 3787 }, { "epoch": 0.48187253530085233, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.85075855255127, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8658974766731262, "num_tokens": 144533989.0, "step": 3788 }, { "epoch": 0.4819997455794428, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.923233032226562, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8527472019195557, "num_tokens": 144577409.0, "step": 3789 }, { "epoch": 0.48212695585803333, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.871033668518066, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8495314121246338, "num_tokens": 144618501.0, "step": 3790 }, { "epoch": 0.48225416613662386, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.84371280670166, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8631844520568848, "num_tokens": 144659454.0, "step": 3791 }, { "epoch": 0.48238137641521434, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.905683517456055, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8625421524047852, "num_tokens": 144695580.0, "step": 3792 }, { "epoch": 0.48250858669380486, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.928918838500977, "learning_rate": 1e-06, "loss": 0.4999, "mean_token_accuracy": 0.840025782585144, "num_tokens": 144735586.0, "step": 3793 }, { "epoch": 0.4826357969723954, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.87021541595459, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8584177494049072, "num_tokens": 144776371.0, "step": 3794 }, { "epoch": 0.48276300725098586, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.922554016113281, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8515832424163818, "num_tokens": 144816724.0, "step": 3795 }, { "epoch": 0.4828902175295764, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.907654762268066, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8634613752365112, "num_tokens": 144854591.0, "step": 3796 }, { "epoch": 0.4830174278081669, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.890118598937988, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.858161211013794, "num_tokens": 144893694.0, "step": 3797 }, { "epoch": 0.4831446380867574, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.919360160827637, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8535606861114502, "num_tokens": 144932985.0, "step": 3798 }, { "epoch": 0.4832718483653479, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.925378799438477, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8566030859947205, "num_tokens": 144967823.0, "step": 3799 }, { "epoch": 0.48339905864393845, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.849268913269043, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8766555190086365, "num_tokens": 145007750.0, "step": 3800 }, { "epoch": 0.4835262689225289, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 9.905900955200195, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8505930304527283, "num_tokens": 145048596.0, "step": 3801 }, { "epoch": 0.48365347920111945, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.959681510925293, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8460187911987305, "num_tokens": 145080868.0, "step": 3802 }, { "epoch": 0.48378068947971, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.924469947814941, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8659807443618774, "num_tokens": 145115865.0, "step": 3803 }, { "epoch": 0.48390789975830045, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.90328598022461, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8701502084732056, "num_tokens": 145151748.0, "step": 3804 }, { "epoch": 0.484035110036891, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.900075912475586, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8570367097854614, "num_tokens": 145187074.0, "step": 3805 }, { "epoch": 0.4841623203154815, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.919952392578125, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8751641511917114, "num_tokens": 145224855.0, "step": 3806 }, { "epoch": 0.484289530594072, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.899797439575195, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8662564754486084, "num_tokens": 145266538.0, "step": 3807 }, { "epoch": 0.4844167408726625, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.878325462341309, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8616667985916138, "num_tokens": 145303560.0, "step": 3808 }, { "epoch": 0.48454395115125304, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.928790092468262, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8581753969192505, "num_tokens": 145345985.0, "step": 3809 }, { "epoch": 0.4846711614298435, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.91996955871582, "learning_rate": 1e-06, "loss": 0.4807, "mean_token_accuracy": 0.8464230895042419, "num_tokens": 145382132.0, "step": 3810 }, { "epoch": 0.48479837170843404, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.924041748046875, "learning_rate": 1e-06, "loss": 0.4737, "mean_token_accuracy": 0.8478928804397583, "num_tokens": 145418992.0, "step": 3811 }, { "epoch": 0.48492558198702457, "ewc_loss": 0.016845703125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.910951614379883, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8564916849136353, "num_tokens": 145455892.0, "step": 3812 }, { "epoch": 0.48505279226561504, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.891203880310059, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8552122116088867, "num_tokens": 145497217.0, "step": 3813 }, { "epoch": 0.48518000254420557, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.935154914855957, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8636982440948486, "num_tokens": 145533822.0, "step": 3814 }, { "epoch": 0.4853072128227961, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.892955780029297, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8774381875991821, "num_tokens": 145572967.0, "step": 3815 }, { "epoch": 0.48543442310138657, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.907045364379883, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8720465898513794, "num_tokens": 145609436.0, "step": 3816 }, { "epoch": 0.4855616333799771, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.96279525756836, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8693259358406067, "num_tokens": 145645083.0, "step": 3817 }, { "epoch": 0.4856888436585676, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.970013618469238, "learning_rate": 1e-06, "loss": 0.5209, "mean_token_accuracy": 0.8386492729187012, "num_tokens": 145680249.0, "step": 3818 }, { "epoch": 0.4858160539371581, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.921865463256836, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8621390461921692, "num_tokens": 145718569.0, "step": 3819 }, { "epoch": 0.4859432642157486, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.963761329650879, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8422845005989075, "num_tokens": 145751868.0, "step": 3820 }, { "epoch": 0.48607047449433916, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.941411018371582, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8620370626449585, "num_tokens": 145787881.0, "step": 3821 }, { "epoch": 0.48619768477292963, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.905710220336914, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8643277883529663, "num_tokens": 145824010.0, "step": 3822 }, { "epoch": 0.48632489505152016, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.928380966186523, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8648160696029663, "num_tokens": 145860981.0, "step": 3823 }, { "epoch": 0.4864521053301107, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.93106460571289, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8620443940162659, "num_tokens": 145900687.0, "step": 3824 }, { "epoch": 0.48657931560870116, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.967079162597656, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8489254713058472, "num_tokens": 145938888.0, "step": 3825 }, { "epoch": 0.4867065258872917, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.965195655822754, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.853076159954071, "num_tokens": 145987060.0, "step": 3826 }, { "epoch": 0.4868337361658822, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.993520736694336, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8528212308883667, "num_tokens": 146026447.0, "step": 3827 }, { "epoch": 0.4869609464444727, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.910235404968262, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8599337339401245, "num_tokens": 146067099.0, "step": 3828 }, { "epoch": 0.4870881567230632, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.984620094299316, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8453452587127686, "num_tokens": 146110767.0, "step": 3829 }, { "epoch": 0.48721536700165374, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.941648483276367, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.849833607673645, "num_tokens": 146147603.0, "step": 3830 }, { "epoch": 0.48734257728024427, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.957796096801758, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8680533170700073, "num_tokens": 146188821.0, "step": 3831 }, { "epoch": 0.48746978755883474, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.928838729858398, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8542963266372681, "num_tokens": 146231886.0, "step": 3832 }, { "epoch": 0.48759699783742527, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.924677848815918, "learning_rate": 1e-06, "loss": 0.5044, "mean_token_accuracy": 0.8348944187164307, "num_tokens": 146271965.0, "step": 3833 }, { "epoch": 0.4877242081160158, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.981939315795898, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8444204926490784, "num_tokens": 146316686.0, "step": 3834 }, { "epoch": 0.4878514183946063, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.907618522644043, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8472564220428467, "num_tokens": 146355098.0, "step": 3835 }, { "epoch": 0.4879786286731968, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.893928527832031, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.863378643989563, "num_tokens": 146390666.0, "step": 3836 }, { "epoch": 0.48810583895178733, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.930253982543945, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8645037412643433, "num_tokens": 146426740.0, "step": 3837 }, { "epoch": 0.4882330492303778, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.933378219604492, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.848176121711731, "num_tokens": 146462668.0, "step": 3838 }, { "epoch": 0.48836025950896833, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.898223876953125, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8504084348678589, "num_tokens": 146500160.0, "step": 3839 }, { "epoch": 0.48848746978755886, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.916747093200684, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8528754115104675, "num_tokens": 146542963.0, "step": 3840 }, { "epoch": 0.48861468006614933, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.890974044799805, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.864046573638916, "num_tokens": 146580649.0, "step": 3841 }, { "epoch": 0.48874189034473986, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.921945571899414, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8643136620521545, "num_tokens": 146620823.0, "step": 3842 }, { "epoch": 0.4888691006233304, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.93403148651123, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8679218888282776, "num_tokens": 146663700.0, "step": 3843 }, { "epoch": 0.48899631090192086, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.888429641723633, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.852225661277771, "num_tokens": 146698211.0, "step": 3844 }, { "epoch": 0.4891235211805114, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.964319229125977, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8556830883026123, "num_tokens": 146738380.0, "step": 3845 }, { "epoch": 0.4892507314591019, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.880008697509766, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8601994514465332, "num_tokens": 146775413.0, "step": 3846 }, { "epoch": 0.4893779417376924, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.926521301269531, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8598100543022156, "num_tokens": 146814774.0, "step": 3847 }, { "epoch": 0.4895051520162829, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.893534660339355, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.867358386516571, "num_tokens": 146853957.0, "step": 3848 }, { "epoch": 0.48963236229487345, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.971158027648926, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8500668406486511, "num_tokens": 146889060.0, "step": 3849 }, { "epoch": 0.4897595725734639, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.908439636230469, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.866976261138916, "num_tokens": 146929849.0, "step": 3850 }, { "epoch": 0.48988678285205445, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.919703483581543, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8598776459693909, "num_tokens": 146968376.0, "step": 3851 }, { "epoch": 0.490013993130645, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.029592514038086, "learning_rate": 1e-06, "loss": 0.4988, "mean_token_accuracy": 0.8418920040130615, "num_tokens": 147002871.0, "step": 3852 }, { "epoch": 0.49014120340923545, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.970044136047363, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.854024350643158, "num_tokens": 147039412.0, "step": 3853 }, { "epoch": 0.490268413687826, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.897171974182129, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8615637421607971, "num_tokens": 147080307.0, "step": 3854 }, { "epoch": 0.4903956239664165, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.977673530578613, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8654430508613586, "num_tokens": 147120131.0, "step": 3855 }, { "epoch": 0.490522834245007, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.95041561126709, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8509863615036011, "num_tokens": 147155760.0, "step": 3856 }, { "epoch": 0.4906500445235975, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.913933753967285, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8621883392333984, "num_tokens": 147195406.0, "step": 3857 }, { "epoch": 0.49077725480218803, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.885026931762695, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8708111047744751, "num_tokens": 147235688.0, "step": 3858 }, { "epoch": 0.4909044650807785, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 10.065974235534668, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8655624389648438, "num_tokens": 147271575.0, "step": 3859 }, { "epoch": 0.49103167535936904, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.929832458496094, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8546968698501587, "num_tokens": 147315473.0, "step": 3860 }, { "epoch": 0.49115888563795956, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.90815544128418, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8615185618400574, "num_tokens": 147359058.0, "step": 3861 }, { "epoch": 0.49128609591655004, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.961594581604004, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8626551032066345, "num_tokens": 147394443.0, "step": 3862 }, { "epoch": 0.49141330619514056, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 9.903264999389648, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8523849248886108, "num_tokens": 147436225.0, "step": 3863 }, { "epoch": 0.4915405164737311, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.948090553283691, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8668745756149292, "num_tokens": 147473449.0, "step": 3864 }, { "epoch": 0.49166772675232157, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.969189643859863, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8560358285903931, "num_tokens": 147510330.0, "step": 3865 }, { "epoch": 0.4917949370309121, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.956843376159668, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8690756559371948, "num_tokens": 147544683.0, "step": 3866 }, { "epoch": 0.4919221473095026, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.907927513122559, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8756084442138672, "num_tokens": 147589330.0, "step": 3867 }, { "epoch": 0.4920493575880931, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.948282241821289, "learning_rate": 1e-06, "loss": 0.5456, "mean_token_accuracy": 0.8267208933830261, "num_tokens": 147626864.0, "step": 3868 }, { "epoch": 0.4921765678666836, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.910675048828125, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.8392593264579773, "num_tokens": 147672006.0, "step": 3869 }, { "epoch": 0.49230377814527415, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.957456588745117, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8485531210899353, "num_tokens": 147717388.0, "step": 3870 }, { "epoch": 0.4924309884238646, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.920483589172363, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8465590476989746, "num_tokens": 147759827.0, "step": 3871 }, { "epoch": 0.49255819870245515, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.002159118652344, "learning_rate": 1e-06, "loss": 0.4893, "mean_token_accuracy": 0.8393851518630981, "num_tokens": 147793742.0, "step": 3872 }, { "epoch": 0.4926854089810457, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.918054580688477, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8679866194725037, "num_tokens": 147829881.0, "step": 3873 }, { "epoch": 0.49281261925963615, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.900525093078613, "learning_rate": 1e-06, "loss": 0.4862, "mean_token_accuracy": 0.8428752422332764, "num_tokens": 147875051.0, "step": 3874 }, { "epoch": 0.4929398295382267, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.93336296081543, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8505919575691223, "num_tokens": 147912786.0, "step": 3875 }, { "epoch": 0.4930670398168172, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.882059097290039, "learning_rate": 1e-06, "loss": 0.5048, "mean_token_accuracy": 0.8380171060562134, "num_tokens": 147954981.0, "step": 3876 }, { "epoch": 0.4931942500954077, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.942707061767578, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8510704040527344, "num_tokens": 147992179.0, "step": 3877 }, { "epoch": 0.4933214603739982, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.858628273010254, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8626932501792908, "num_tokens": 148029866.0, "step": 3878 }, { "epoch": 0.49344867065258874, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.948548316955566, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8574728965759277, "num_tokens": 148070057.0, "step": 3879 }, { "epoch": 0.4935758809311792, "ewc_loss": 0.017333984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.971858024597168, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.861742377281189, "num_tokens": 148108461.0, "step": 3880 }, { "epoch": 0.49370309120976974, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.916620254516602, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8506158590316772, "num_tokens": 148147609.0, "step": 3881 }, { "epoch": 0.49383030148836027, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.945893287658691, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8521015644073486, "num_tokens": 148190475.0, "step": 3882 }, { "epoch": 0.4939575117669508, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.984009742736816, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8689405918121338, "num_tokens": 148221187.0, "step": 3883 }, { "epoch": 0.49408472204554127, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.92452621459961, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8662937879562378, "num_tokens": 148260916.0, "step": 3884 }, { "epoch": 0.4942119323241318, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.946342468261719, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8631620407104492, "num_tokens": 148295591.0, "step": 3885 }, { "epoch": 0.4943391426027223, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.922247886657715, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.853157103061676, "num_tokens": 148333445.0, "step": 3886 }, { "epoch": 0.4944663528813128, "ewc_loss": 0.017333984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.956144332885742, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8662616610527039, "num_tokens": 148367951.0, "step": 3887 }, { "epoch": 0.4945935631599033, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.928397178649902, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8491191864013672, "num_tokens": 148408491.0, "step": 3888 }, { "epoch": 0.49472077343849385, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.92622184753418, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8571842908859253, "num_tokens": 148443883.0, "step": 3889 }, { "epoch": 0.4948479837170843, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.975274085998535, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8591145277023315, "num_tokens": 148480511.0, "step": 3890 }, { "epoch": 0.49497519399567486, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.94256591796875, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8526009917259216, "num_tokens": 148514759.0, "step": 3891 }, { "epoch": 0.4951024042742654, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.944604873657227, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8596800565719604, "num_tokens": 148553209.0, "step": 3892 }, { "epoch": 0.49522961455285586, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.940705299377441, "learning_rate": 1e-06, "loss": 0.4649, "mean_token_accuracy": 0.8486221432685852, "num_tokens": 148590387.0, "step": 3893 }, { "epoch": 0.4953568248314464, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.965700149536133, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8551684617996216, "num_tokens": 148625201.0, "step": 3894 }, { "epoch": 0.4954840351100369, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.940494537353516, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8600447773933411, "num_tokens": 148665989.0, "step": 3895 }, { "epoch": 0.4956112453886274, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.928555488586426, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8600316047668457, "num_tokens": 148704839.0, "step": 3896 }, { "epoch": 0.4957384556672179, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.9692964553833, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8676885366439819, "num_tokens": 148740523.0, "step": 3897 }, { "epoch": 0.49586566594580844, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.929173469543457, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8728849291801453, "num_tokens": 148779103.0, "step": 3898 }, { "epoch": 0.4959928762243989, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.961087226867676, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8686408996582031, "num_tokens": 148814659.0, "step": 3899 }, { "epoch": 0.49612008650298944, "ewc_loss": 0.01708984375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.908905982971191, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8779033422470093, "num_tokens": 148856864.0, "step": 3900 }, { "epoch": 0.49624729678157997, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.933173179626465, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8638839721679688, "num_tokens": 148890309.0, "step": 3901 }, { "epoch": 0.49637450706017044, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.988883018493652, "learning_rate": 1e-06, "loss": 0.4924, "mean_token_accuracy": 0.8426697254180908, "num_tokens": 148926662.0, "step": 3902 }, { "epoch": 0.496501717338761, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.943931579589844, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.845039963722229, "num_tokens": 148963032.0, "step": 3903 }, { "epoch": 0.4966289276173515, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.985403060913086, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8639549016952515, "num_tokens": 148996684.0, "step": 3904 }, { "epoch": 0.496756137895942, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 9.937295913696289, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8575469851493835, "num_tokens": 149032735.0, "step": 3905 }, { "epoch": 0.4968833481745325, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.923361778259277, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8603225946426392, "num_tokens": 149069415.0, "step": 3906 }, { "epoch": 0.49701055845312303, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.95906925201416, "learning_rate": 1e-06, "loss": 0.4915, "mean_token_accuracy": 0.8437016010284424, "num_tokens": 149115042.0, "step": 3907 }, { "epoch": 0.4971377687317135, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.020920753479004, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8655960559844971, "num_tokens": 149149767.0, "step": 3908 }, { "epoch": 0.49726497901030403, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.943984031677246, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8510712385177612, "num_tokens": 149193000.0, "step": 3909 }, { "epoch": 0.49739218928889456, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.926759719848633, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8548892140388489, "num_tokens": 149226963.0, "step": 3910 }, { "epoch": 0.49751939956748503, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.971578598022461, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8528563380241394, "num_tokens": 149268139.0, "step": 3911 }, { "epoch": 0.49764660984607556, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.957154273986816, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8653564453125, "num_tokens": 149305882.0, "step": 3912 }, { "epoch": 0.4977738201246661, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.960898399353027, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8548057675361633, "num_tokens": 149344119.0, "step": 3913 }, { "epoch": 0.49790103040325656, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.971552848815918, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8485783338546753, "num_tokens": 149384102.0, "step": 3914 }, { "epoch": 0.4980282406818471, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.908855438232422, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8554952144622803, "num_tokens": 149425677.0, "step": 3915 }, { "epoch": 0.4981554509604376, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.997858047485352, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8539276123046875, "num_tokens": 149463275.0, "step": 3916 }, { "epoch": 0.4982826612390281, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.967880249023438, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8724029064178467, "num_tokens": 149503950.0, "step": 3917 }, { "epoch": 0.4984098715176186, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.006492614746094, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8487871289253235, "num_tokens": 149541192.0, "step": 3918 }, { "epoch": 0.49853708179620915, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.001602172851562, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8547204732894897, "num_tokens": 149581784.0, "step": 3919 }, { "epoch": 0.4986642920747996, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.96908950805664, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8633646965026855, "num_tokens": 149619312.0, "step": 3920 }, { "epoch": 0.49879150235339015, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.94328784942627, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8548929691314697, "num_tokens": 149657443.0, "step": 3921 }, { "epoch": 0.4989187126319807, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.045927047729492, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8690624833106995, "num_tokens": 149694992.0, "step": 3922 }, { "epoch": 0.49904592291057115, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.98441219329834, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8540226221084595, "num_tokens": 149729554.0, "step": 3923 }, { "epoch": 0.4991731331891617, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.920964241027832, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8632259964942932, "num_tokens": 149770051.0, "step": 3924 }, { "epoch": 0.4993003434677522, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.97736930847168, "learning_rate": 1e-06, "loss": 0.5396, "mean_token_accuracy": 0.8264257311820984, "num_tokens": 149806724.0, "step": 3925 }, { "epoch": 0.4994275537463427, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.936236381530762, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8492674827575684, "num_tokens": 149849874.0, "step": 3926 }, { "epoch": 0.4995547640249332, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.999342918395996, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8570135831832886, "num_tokens": 149888076.0, "step": 3927 }, { "epoch": 0.49968197430352373, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.970992088317871, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8582543134689331, "num_tokens": 149923750.0, "step": 3928 }, { "epoch": 0.4998091845821142, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.969247817993164, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8448883295059204, "num_tokens": 149960657.0, "step": 3929 }, { "epoch": 0.49993639486070474, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.01842975616455, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8733656406402588, "num_tokens": 149995983.0, "step": 3930 }, { "epoch": 0.5000636051392953, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.955659866333008, "learning_rate": 1e-06, "loss": 0.5059, "mean_token_accuracy": 0.8350815176963806, "num_tokens": 150037508.0, "step": 3931 }, { "epoch": 0.5001908154178858, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.999573707580566, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8539045453071594, "num_tokens": 150077806.0, "step": 3932 }, { "epoch": 0.5003180256964763, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.97901725769043, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8561161756515503, "num_tokens": 150117589.0, "step": 3933 }, { "epoch": 0.5004452359750667, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.999195098876953, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8501114845275879, "num_tokens": 150161579.0, "step": 3934 }, { "epoch": 0.5005724462536573, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.004744529724121, "learning_rate": 1e-06, "loss": 0.4904, "mean_token_accuracy": 0.8445098400115967, "num_tokens": 150208018.0, "step": 3935 }, { "epoch": 0.5006996565322478, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.97346305847168, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8590226173400879, "num_tokens": 150250164.0, "step": 3936 }, { "epoch": 0.5008268668108383, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.987812042236328, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8758907914161682, "num_tokens": 150289972.0, "step": 3937 }, { "epoch": 0.5009540770894289, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.96707534790039, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8553288578987122, "num_tokens": 150333656.0, "step": 3938 }, { "epoch": 0.5010812873680194, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.018149375915527, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8704061508178711, "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": 10.013349533081055, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8603587746620178, "num_tokens": 150407472.0, "step": 3940 }, { "epoch": 0.5013357079252003, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 9.987486839294434, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8634839057922363, "num_tokens": 150444560.0, "step": 3941 }, { "epoch": 0.5014629182037909, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.929651260375977, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.861402153968811, "num_tokens": 150483621.0, "step": 3942 }, { "epoch": 0.5015901284823814, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.05266284942627, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8576193451881409, "num_tokens": 150520414.0, "step": 3943 }, { "epoch": 0.5017173387609719, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.987475395202637, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8486751317977905, "num_tokens": 150556589.0, "step": 3944 }, { "epoch": 0.5018445490395624, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.957878112792969, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8559876084327698, "num_tokens": 150592393.0, "step": 3945 }, { "epoch": 0.5019717593181529, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.051977157592773, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8443098068237305, "num_tokens": 150630457.0, "step": 3946 }, { "epoch": 0.5020989695967434, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.970333099365234, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8645867109298706, "num_tokens": 150659482.0, "step": 3947 }, { "epoch": 0.5022261798753339, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.99831485748291, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8568660616874695, "num_tokens": 150700920.0, "step": 3948 }, { "epoch": 0.5023533901539244, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.033341407775879, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.868331789970398, "num_tokens": 150736393.0, "step": 3949 }, { "epoch": 0.502480600432515, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.968767166137695, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.854106605052948, "num_tokens": 150783848.0, "step": 3950 }, { "epoch": 0.5026078107111055, "ewc_loss": 0.01708984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.93008041381836, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8545689582824707, "num_tokens": 150824179.0, "step": 3951 }, { "epoch": 0.5027350209896959, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.056151390075684, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8576576709747314, "num_tokens": 150862775.0, "step": 3952 }, { "epoch": 0.5028622312682864, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.95757007598877, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8566967844963074, "num_tokens": 150900005.0, "step": 3953 }, { "epoch": 0.502989441546877, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.006525993347168, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8539648652076721, "num_tokens": 150936148.0, "step": 3954 }, { "epoch": 0.5031166518254675, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.99101448059082, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8544889688491821, "num_tokens": 150971661.0, "step": 3955 }, { "epoch": 0.503243862104058, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.003315925598145, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8680692911148071, "num_tokens": 151009521.0, "step": 3956 }, { "epoch": 0.5033710723826486, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.959179878234863, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8541107773780823, "num_tokens": 151053641.0, "step": 3957 }, { "epoch": 0.5034982826612391, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.968057632446289, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8610534071922302, "num_tokens": 151094476.0, "step": 3958 }, { "epoch": 0.5036254929398295, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.01717472076416, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8509645462036133, "num_tokens": 151126381.0, "step": 3959 }, { "epoch": 0.50375270321842, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.010663986206055, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8618192076683044, "num_tokens": 151156875.0, "step": 3960 }, { "epoch": 0.5038799134970106, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.957734107971191, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8534122705459595, "num_tokens": 151194394.0, "step": 3961 }, { "epoch": 0.5040071237756011, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.952582359313965, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8438001275062561, "num_tokens": 151234472.0, "step": 3962 }, { "epoch": 0.5041343340541916, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 10.051372528076172, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8529322147369385, "num_tokens": 151270426.0, "step": 3963 }, { "epoch": 0.5042615443327821, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.9679594039917, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8552268743515015, "num_tokens": 151305793.0, "step": 3964 }, { "epoch": 0.5043887546113726, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.98682689666748, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8453168869018555, "num_tokens": 151345116.0, "step": 3965 }, { "epoch": 0.5045159648899631, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.959403991699219, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8620372414588928, "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": 9.997965812683105, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8586514592170715, "num_tokens": 151426884.0, "step": 3967 }, { "epoch": 0.5047703854471441, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.973313331604004, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.849570095539093, "num_tokens": 151467596.0, "step": 3968 }, { "epoch": 0.5048975957257347, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.966276168823242, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8766634464263916, "num_tokens": 151502131.0, "step": 3969 }, { "epoch": 0.5050248060043252, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.00716781616211, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8542532920837402, "num_tokens": 151538932.0, "step": 3970 }, { "epoch": 0.5051520162829156, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.963872909545898, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8564997315406799, "num_tokens": 151577839.0, "step": 3971 }, { "epoch": 0.5052792265615061, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.969009399414062, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8569049835205078, "num_tokens": 151618984.0, "step": 3972 }, { "epoch": 0.5054064368400967, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.992523193359375, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.856971263885498, "num_tokens": 151658392.0, "step": 3973 }, { "epoch": 0.5055336471186872, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.03397274017334, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8672659397125244, "num_tokens": 151689781.0, "step": 3974 }, { "epoch": 0.5056608573972777, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.044806480407715, "learning_rate": 1e-06, "loss": 0.5007, "mean_token_accuracy": 0.8432343006134033, "num_tokens": 151729612.0, "step": 3975 }, { "epoch": 0.5057880676758683, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.027454376220703, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8441305160522461, "num_tokens": 151769270.0, "step": 3976 }, { "epoch": 0.5059152779544587, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.042860984802246, "learning_rate": 1e-06, "loss": 0.4926, "mean_token_accuracy": 0.8383074402809143, "num_tokens": 151803936.0, "step": 3977 }, { "epoch": 0.5060424882330492, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.033130645751953, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8451783657073975, "num_tokens": 151840429.0, "step": 3978 }, { "epoch": 0.5061696985116397, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.967894554138184, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8583083748817444, "num_tokens": 151883296.0, "step": 3979 }, { "epoch": 0.5062969087902303, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.976521492004395, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8722016215324402, "num_tokens": 151917658.0, "step": 3980 }, { "epoch": 0.5064241190688208, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.01809310913086, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8564302921295166, "num_tokens": 151958822.0, "step": 3981 }, { "epoch": 0.5065513293474113, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.977051734924316, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8518286347389221, "num_tokens": 151994964.0, "step": 3982 }, { "epoch": 0.5066785396260017, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.965452194213867, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8518623113632202, "num_tokens": 152037013.0, "step": 3983 }, { "epoch": 0.5068057499045923, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.95827865600586, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8530205488204956, "num_tokens": 152074821.0, "step": 3984 }, { "epoch": 0.5069329601831828, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.012901306152344, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8494968414306641, "num_tokens": 152116667.0, "step": 3985 }, { "epoch": 0.5070601704617733, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.957736015319824, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8609252572059631, "num_tokens": 152152714.0, "step": 3986 }, { "epoch": 0.5071873807403638, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.988689422607422, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8427018523216248, "num_tokens": 152190230.0, "step": 3987 }, { "epoch": 0.5073145910189544, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.982259750366211, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8560569286346436, "num_tokens": 152228887.0, "step": 3988 }, { "epoch": 0.5074418012975448, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.98598575592041, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8655449748039246, "num_tokens": 152271863.0, "step": 3989 }, { "epoch": 0.5075690115761353, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.96682071685791, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8517976999282837, "num_tokens": 152310765.0, "step": 3990 }, { "epoch": 0.5076962218547258, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.994778633117676, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8793193697929382, "num_tokens": 152342545.0, "step": 3991 }, { "epoch": 0.5078234321333164, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.005414962768555, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8409693241119385, "num_tokens": 152386223.0, "step": 3992 }, { "epoch": 0.5079506424119069, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.976726531982422, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.86351478099823, "num_tokens": 152421811.0, "step": 3993 }, { "epoch": 0.5080778526904974, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.047420501708984, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8448024988174438, "num_tokens": 152454581.0, "step": 3994 }, { "epoch": 0.5082050629690879, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.962079048156738, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8474195599555969, "num_tokens": 152490591.0, "step": 3995 }, { "epoch": 0.5083322732476784, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.942306518554688, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8795045614242554, "num_tokens": 152532290.0, "step": 3996 }, { "epoch": 0.5084594835262689, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.004456520080566, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8623875379562378, "num_tokens": 152571984.0, "step": 3997 }, { "epoch": 0.5085866938048594, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.014663696289062, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8567962646484375, "num_tokens": 152611082.0, "step": 3998 }, { "epoch": 0.50871390408345, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.9515962600708, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8645821809768677, "num_tokens": 152655303.0, "step": 3999 }, { "epoch": 0.5088411143620405, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.025619506835938, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8559530973434448, "num_tokens": 152688432.0, "step": 4000 }, { "epoch": 0.5089683246406309, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.959331512451172, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8530924916267395, "num_tokens": 152725637.0, "step": 4001 }, { "epoch": 0.5090955349192214, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.011731147766113, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8544862270355225, "num_tokens": 152763108.0, "step": 4002 }, { "epoch": 0.509222745197812, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.974082946777344, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.863006591796875, "num_tokens": 152800659.0, "step": 4003 }, { "epoch": 0.5093499554764025, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.94912338256836, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8476234674453735, "num_tokens": 152842797.0, "step": 4004 }, { "epoch": 0.509477165754993, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.971200942993164, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.848468542098999, "num_tokens": 152888257.0, "step": 4005 }, { "epoch": 0.5096043760335836, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.986042976379395, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8558088541030884, "num_tokens": 152929089.0, "step": 4006 }, { "epoch": 0.5097315863121741, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.976740837097168, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8508052825927734, "num_tokens": 152972710.0, "step": 4007 }, { "epoch": 0.5098587965907645, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.97553825378418, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8568167686462402, "num_tokens": 153008867.0, "step": 4008 }, { "epoch": 0.509986006869355, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.976408004760742, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8517321944236755, "num_tokens": 153051497.0, "step": 4009 }, { "epoch": 0.5101132171479456, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.012022018432617, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.872678279876709, "num_tokens": 153089295.0, "step": 4010 }, { "epoch": 0.5102404274265361, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.998591423034668, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8615704774856567, "num_tokens": 153123957.0, "step": 4011 }, { "epoch": 0.5103676377051266, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.988874435424805, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8537896275520325, "num_tokens": 153165248.0, "step": 4012 }, { "epoch": 0.5104948479837171, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 10.01691722869873, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8717976212501526, "num_tokens": 153206163.0, "step": 4013 }, { "epoch": 0.5106220582623076, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.978076934814453, "learning_rate": 1e-06, "loss": 0.4735, "mean_token_accuracy": 0.8479735851287842, "num_tokens": 153248666.0, "step": 4014 }, { "epoch": 0.5107492685408981, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.001214981079102, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8802827596664429, "num_tokens": 153289009.0, "step": 4015 }, { "epoch": 0.5108764788194886, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.06442642211914, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8625269532203674, "num_tokens": 153321354.0, "step": 4016 }, { "epoch": 0.5110036890980791, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.973224639892578, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8570479154586792, "num_tokens": 153361061.0, "step": 4017 }, { "epoch": 0.5111308993766697, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.049118995666504, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8509916067123413, "num_tokens": 153397002.0, "step": 4018 }, { "epoch": 0.5112581096552602, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.019211769104004, "learning_rate": 1e-06, "loss": 0.4849, "mean_token_accuracy": 0.8436617851257324, "num_tokens": 153435264.0, "step": 4019 }, { "epoch": 0.5113853199338506, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.985394477844238, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8551873564720154, "num_tokens": 153474751.0, "step": 4020 }, { "epoch": 0.5115125302124411, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.981003761291504, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8657434582710266, "num_tokens": 153510084.0, "step": 4021 }, { "epoch": 0.5116397404910317, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.013068199157715, "learning_rate": 1e-06, "loss": 0.4905, "mean_token_accuracy": 0.8442333936691284, "num_tokens": 153550387.0, "step": 4022 }, { "epoch": 0.5117669507696222, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.976773262023926, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8789042234420776, "num_tokens": 153592441.0, "step": 4023 }, { "epoch": 0.5118941610482127, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.065048217773438, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8642648458480835, "num_tokens": 153629202.0, "step": 4024 }, { "epoch": 0.5120213713268033, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.992271423339844, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8523737788200378, "num_tokens": 153669243.0, "step": 4025 }, { "epoch": 0.5121485816053937, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.058345794677734, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8518590927124023, "num_tokens": 153709206.0, "step": 4026 }, { "epoch": 0.5122757918839842, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.075008392333984, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8636088371276855, "num_tokens": 153741251.0, "step": 4027 }, { "epoch": 0.5124030021625747, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.031855583190918, "learning_rate": 1e-06, "loss": 0.4933, "mean_token_accuracy": 0.843254804611206, "num_tokens": 153775559.0, "step": 4028 }, { "epoch": 0.5125302124411653, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.030070304870605, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8613328337669373, "num_tokens": 153819158.0, "step": 4029 }, { "epoch": 0.5126574227197558, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.028474807739258, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8508237600326538, "num_tokens": 153856231.0, "step": 4030 }, { "epoch": 0.5127846329983463, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.00581169128418, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8674517869949341, "num_tokens": 153897125.0, "step": 4031 }, { "epoch": 0.5129118432769367, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.07905387878418, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8639676570892334, "num_tokens": 153933552.0, "step": 4032 }, { "epoch": 0.5130390535555273, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.01869010925293, "learning_rate": 1e-06, "loss": 0.4971, "mean_token_accuracy": 0.8428528904914856, "num_tokens": 153974337.0, "step": 4033 }, { "epoch": 0.5131662638341178, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.005512237548828, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8520898818969727, "num_tokens": 154018911.0, "step": 4034 }, { "epoch": 0.5132934741127083, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.007352828979492, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8673853874206543, "num_tokens": 154056441.0, "step": 4035 }, { "epoch": 0.5134206843912988, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.050326347351074, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.847986102104187, "num_tokens": 154096326.0, "step": 4036 }, { "epoch": 0.5135478946698894, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.031662940979004, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8609068989753723, "num_tokens": 154137223.0, "step": 4037 }, { "epoch": 0.5136751049484798, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.015796661376953, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8551035523414612, "num_tokens": 154168312.0, "step": 4038 }, { "epoch": 0.5138023152270703, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.030122756958008, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8698983788490295, "num_tokens": 154203125.0, "step": 4039 }, { "epoch": 0.5139295255056608, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.971893310546875, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8511126041412354, "num_tokens": 154248437.0, "step": 4040 }, { "epoch": 0.5140567357842514, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.026626586914062, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8611465096473694, "num_tokens": 154286303.0, "step": 4041 }, { "epoch": 0.5141839460628419, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.989742279052734, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8475489020347595, "num_tokens": 154328638.0, "step": 4042 }, { "epoch": 0.5143111563414324, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.041916847229004, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.856312096118927, "num_tokens": 154362953.0, "step": 4043 }, { "epoch": 0.5144383666200228, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.057470321655273, "learning_rate": 1e-06, "loss": 0.4899, "mean_token_accuracy": 0.8454592227935791, "num_tokens": 154401289.0, "step": 4044 }, { "epoch": 0.5145655768986134, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.979595184326172, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8612646460533142, "num_tokens": 154438779.0, "step": 4045 }, { "epoch": 0.5146927871772039, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.083048820495605, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8510057926177979, "num_tokens": 154467032.0, "step": 4046 }, { "epoch": 0.5148199974557944, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.030714988708496, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8663439750671387, "num_tokens": 154505222.0, "step": 4047 }, { "epoch": 0.514947207734385, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.026750564575195, "learning_rate": 1e-06, "loss": 0.5001, "mean_token_accuracy": 0.8440752029418945, "num_tokens": 154544953.0, "step": 4048 }, { "epoch": 0.5150744180129755, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.020938873291016, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8693000078201294, "num_tokens": 154586193.0, "step": 4049 }, { "epoch": 0.5152016282915659, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.990865707397461, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8603031635284424, "num_tokens": 154629531.0, "step": 4050 }, { "epoch": 0.5153288385701564, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.057218551635742, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8716170787811279, "num_tokens": 154659482.0, "step": 4051 }, { "epoch": 0.515456048848747, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.06994342803955, "learning_rate": 1e-06, "loss": 0.5391, "mean_token_accuracy": 0.8244861364364624, "num_tokens": 154702645.0, "step": 4052 }, { "epoch": 0.5155832591273375, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.013932228088379, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8654402494430542, "num_tokens": 154743329.0, "step": 4053 }, { "epoch": 0.515710469405928, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 9.985318183898926, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8537076711654663, "num_tokens": 154783460.0, "step": 4054 }, { "epoch": 0.5158376796845185, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.030238151550293, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8509489297866821, "num_tokens": 154827209.0, "step": 4055 }, { "epoch": 0.5159648899631091, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.034468650817871, "learning_rate": 1e-06, "loss": 0.5209, "mean_token_accuracy": 0.8321066498756409, "num_tokens": 154866577.0, "step": 4056 }, { "epoch": 0.5160921002416995, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.010817527770996, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8592930436134338, "num_tokens": 154912191.0, "step": 4057 }, { "epoch": 0.51621931052029, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.002434730529785, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8440966606140137, "num_tokens": 154953412.0, "step": 4058 }, { "epoch": 0.5163465207988805, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.001045227050781, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8685751557350159, "num_tokens": 154989913.0, "step": 4059 }, { "epoch": 0.5164737310774711, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.016167640686035, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8444386124610901, "num_tokens": 155038094.0, "step": 4060 }, { "epoch": 0.5166009413560616, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.0077543258667, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8469159603118896, "num_tokens": 155075900.0, "step": 4061 }, { "epoch": 0.5167281516346521, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.992424964904785, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8559813499450684, "num_tokens": 155117038.0, "step": 4062 }, { "epoch": 0.5168553619132426, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.06178092956543, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8709284067153931, "num_tokens": 155151442.0, "step": 4063 }, { "epoch": 0.5169825721918331, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.019105911254883, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8645886182785034, "num_tokens": 155192288.0, "step": 4064 }, { "epoch": 0.5171097824704236, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.95851993560791, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8594480156898499, "num_tokens": 155238931.0, "step": 4065 }, { "epoch": 0.5172369927490141, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.01612663269043, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8698452115058899, "num_tokens": 155276644.0, "step": 4066 }, { "epoch": 0.5173642030276047, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.054783821105957, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8690100312232971, "num_tokens": 155307638.0, "step": 4067 }, { "epoch": 0.5174914133061952, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.014443397521973, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8502131700515747, "num_tokens": 155343255.0, "step": 4068 }, { "epoch": 0.5176186235847856, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.01920223236084, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8587490320205688, "num_tokens": 155382283.0, "step": 4069 }, { "epoch": 0.5177458338633761, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.034195899963379, "learning_rate": 1e-06, "loss": 0.5125, "mean_token_accuracy": 0.8345487713813782, "num_tokens": 155427096.0, "step": 4070 }, { "epoch": 0.5178730441419667, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 9.98476505279541, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8453893065452576, "num_tokens": 155467472.0, "step": 4071 }, { "epoch": 0.5180002544205572, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.993499755859375, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8513646125793457, "num_tokens": 155508350.0, "step": 4072 }, { "epoch": 0.5181274646991477, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 9.995134353637695, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8802204132080078, "num_tokens": 155547215.0, "step": 4073 }, { "epoch": 0.5182546749777382, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 9.997454643249512, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8618137836456299, "num_tokens": 155586658.0, "step": 4074 }, { "epoch": 0.5183818852563287, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.022111892700195, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8590356111526489, "num_tokens": 155620944.0, "step": 4075 }, { "epoch": 0.5185090955349192, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.065958023071289, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8545364141464233, "num_tokens": 155662699.0, "step": 4076 }, { "epoch": 0.5186363058135097, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.030824661254883, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8444330096244812, "num_tokens": 155707783.0, "step": 4077 }, { "epoch": 0.5187635160921003, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.03847885131836, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8629050254821777, "num_tokens": 155745227.0, "step": 4078 }, { "epoch": 0.5188907263706908, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 9.994194030761719, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8583608865737915, "num_tokens": 155786466.0, "step": 4079 }, { "epoch": 0.5190179366492813, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.024840354919434, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.858535885810852, "num_tokens": 155833145.0, "step": 4080 }, { "epoch": 0.5191451469278717, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.086617469787598, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8477060198783875, "num_tokens": 155869533.0, "step": 4081 }, { "epoch": 0.5192723572064623, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.020548820495605, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8562080264091492, "num_tokens": 155909329.0, "step": 4082 }, { "epoch": 0.5193995674850528, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.045927047729492, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8655984401702881, "num_tokens": 155947848.0, "step": 4083 }, { "epoch": 0.5195267777636433, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.011018753051758, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8614847660064697, "num_tokens": 155982693.0, "step": 4084 }, { "epoch": 0.5196539880422338, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.991233825683594, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8559701442718506, "num_tokens": 156023414.0, "step": 4085 }, { "epoch": 0.5197811983208244, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.039710998535156, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8569773435592651, "num_tokens": 156063488.0, "step": 4086 }, { "epoch": 0.5199084085994148, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.015390396118164, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.856952428817749, "num_tokens": 156103081.0, "step": 4087 }, { "epoch": 0.5200356188780053, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.06767749786377, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.848642110824585, "num_tokens": 156139453.0, "step": 4088 }, { "epoch": 0.5201628291565958, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.017794609069824, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8514635562896729, "num_tokens": 156176564.0, "step": 4089 }, { "epoch": 0.5202900394351864, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.980142593383789, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8596619367599487, "num_tokens": 156221278.0, "step": 4090 }, { "epoch": 0.5204172497137769, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.064496040344238, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.862182080745697, "num_tokens": 156259124.0, "step": 4091 }, { "epoch": 0.5205444599923674, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.061616897583008, "learning_rate": 1e-06, "loss": 0.4842, "mean_token_accuracy": 0.8471639156341553, "num_tokens": 156297832.0, "step": 4092 }, { "epoch": 0.5206716702709578, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.062410354614258, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8629601001739502, "num_tokens": 156329745.0, "step": 4093 }, { "epoch": 0.5207988805495484, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.045557022094727, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8573799133300781, "num_tokens": 156360154.0, "step": 4094 }, { "epoch": 0.5209260908281389, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.051547050476074, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8601769804954529, "num_tokens": 156399360.0, "step": 4095 }, { "epoch": 0.5210533011067294, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.028987884521484, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.856231689453125, "num_tokens": 156438782.0, "step": 4096 }, { "epoch": 0.52118051138532, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.057111740112305, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8627798557281494, "num_tokens": 156470261.0, "step": 4097 }, { "epoch": 0.5213077216639105, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.049410820007324, "learning_rate": 1e-06, "loss": 0.4992, "mean_token_accuracy": 0.8391120433807373, "num_tokens": 156510493.0, "step": 4098 }, { "epoch": 0.5214349319425009, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.071100234985352, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8593491911888123, "num_tokens": 156545040.0, "step": 4099 }, { "epoch": 0.5215621422210914, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.054618835449219, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8551279902458191, "num_tokens": 156583474.0, "step": 4100 }, { "epoch": 0.521689352499682, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.07151985168457, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8572083711624146, "num_tokens": 156627137.0, "step": 4101 }, { "epoch": 0.5218165627782725, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.030842781066895, "learning_rate": 1e-06, "loss": 0.4804, "mean_token_accuracy": 0.8456452488899231, "num_tokens": 156664726.0, "step": 4102 }, { "epoch": 0.521943773056863, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.060901641845703, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8745319247245789, "num_tokens": 156707014.0, "step": 4103 }, { "epoch": 0.5220709833354535, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.060439109802246, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8538203239440918, "num_tokens": 156747784.0, "step": 4104 }, { "epoch": 0.522198193614044, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.163756370544434, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8558860421180725, "num_tokens": 156784760.0, "step": 4105 }, { "epoch": 0.5223254038926345, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.056730270385742, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8542734980583191, "num_tokens": 156822899.0, "step": 4106 }, { "epoch": 0.522452614171225, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.010573387145996, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8635751008987427, "num_tokens": 156861308.0, "step": 4107 }, { "epoch": 0.5225798244498155, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.10445785522461, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.864334762096405, "num_tokens": 156901981.0, "step": 4108 }, { "epoch": 0.5227070347284061, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.042709350585938, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8623360395431519, "num_tokens": 156942215.0, "step": 4109 }, { "epoch": 0.5228342450069966, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.659614562988281, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8637060523033142, "num_tokens": 156983820.0, "step": 4110 }, { "epoch": 0.5229614552855871, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.055846214294434, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8697648048400879, "num_tokens": 157016215.0, "step": 4111 }, { "epoch": 0.5230886655641775, "ewc_loss": 0.017333984375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 9.959616661071777, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8575552701950073, "num_tokens": 157052071.0, "step": 4112 }, { "epoch": 0.5232158758427681, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.097794532775879, "learning_rate": 1e-06, "loss": 0.5152, "mean_token_accuracy": 0.8357182741165161, "num_tokens": 157097711.0, "step": 4113 }, { "epoch": 0.5233430861213586, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 9.990521430969238, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8562642335891724, "num_tokens": 157134801.0, "step": 4114 }, { "epoch": 0.5234702963999491, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.06613826751709, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8577540516853333, "num_tokens": 157171885.0, "step": 4115 }, { "epoch": 0.5235975066785397, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.030675888061523, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8560112714767456, "num_tokens": 157211915.0, "step": 4116 }, { "epoch": 0.5237247169571302, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.101465225219727, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8512250781059265, "num_tokens": 157260564.0, "step": 4117 }, { "epoch": 0.5238519272357206, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.028064727783203, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8591322898864746, "num_tokens": 157294676.0, "step": 4118 }, { "epoch": 0.5239791375143111, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.053569793701172, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8677471876144409, "num_tokens": 157334607.0, "step": 4119 }, { "epoch": 0.5241063477929017, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.121596336364746, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8588641285896301, "num_tokens": 157372021.0, "step": 4120 }, { "epoch": 0.5242335580714922, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.04299545288086, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8620550632476807, "num_tokens": 157410092.0, "step": 4121 }, { "epoch": 0.5243607683500827, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.047046661376953, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8542042374610901, "num_tokens": 157447979.0, "step": 4122 }, { "epoch": 0.5244879786286732, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.073492050170898, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8533664345741272, "num_tokens": 157481376.0, "step": 4123 }, { "epoch": 0.5246151889072637, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.118697166442871, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8637304306030273, "num_tokens": 157519555.0, "step": 4124 }, { "epoch": 0.5247423991858542, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.072407722473145, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8562471866607666, "num_tokens": 157551849.0, "step": 4125 }, { "epoch": 0.5248696094644447, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.063055992126465, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8570756316184998, "num_tokens": 157590823.0, "step": 4126 }, { "epoch": 0.5249968197430352, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.059762001037598, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.867823600769043, "num_tokens": 157633771.0, "step": 4127 }, { "epoch": 0.5251240300216258, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.063789367675781, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8709684610366821, "num_tokens": 157667796.0, "step": 4128 }, { "epoch": 0.5252512403002163, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.066705703735352, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8679423332214355, "num_tokens": 157709452.0, "step": 4129 }, { "epoch": 0.5253784505788067, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.070796966552734, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8495895266532898, "num_tokens": 157741407.0, "step": 4130 }, { "epoch": 0.5255056608573972, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.096755981445312, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8681314587593079, "num_tokens": 157777693.0, "step": 4131 }, { "epoch": 0.5256328711359878, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.024198532104492, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8559603691101074, "num_tokens": 157814729.0, "step": 4132 }, { "epoch": 0.5257600814145783, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.050143241882324, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8675937652587891, "num_tokens": 157853698.0, "step": 4133 }, { "epoch": 0.5258872916931688, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.054572105407715, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8625443577766418, "num_tokens": 157888580.0, "step": 4134 }, { "epoch": 0.5260145019717594, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 9.981112480163574, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8555802702903748, "num_tokens": 157928434.0, "step": 4135 }, { "epoch": 0.5261417122503498, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.090415000915527, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8530327081680298, "num_tokens": 157968936.0, "step": 4136 }, { "epoch": 0.5262689225289403, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.08345890045166, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8467848300933838, "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": 10.033759117126465, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8655620813369751, "num_tokens": 158047569.0, "step": 4138 }, { "epoch": 0.5265233430861214, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.043432235717773, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8563381433486938, "num_tokens": 158085733.0, "step": 4139 }, { "epoch": 0.5266505533647119, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.08043098449707, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8476518988609314, "num_tokens": 158123906.0, "step": 4140 }, { "epoch": 0.5267777636433024, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.002633094787598, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8642048835754395, "num_tokens": 158163663.0, "step": 4141 }, { "epoch": 0.5269049739218928, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 10.057796478271484, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8455674648284912, "num_tokens": 158205878.0, "step": 4142 }, { "epoch": 0.5270321842004834, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.000077247619629, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8611609935760498, "num_tokens": 158241608.0, "step": 4143 }, { "epoch": 0.5271593944790739, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.058426856994629, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8700251579284668, "num_tokens": 158274259.0, "step": 4144 }, { "epoch": 0.5272866047576644, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.108304977416992, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8643822073936462, "num_tokens": 158304793.0, "step": 4145 }, { "epoch": 0.527413815036255, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.03480052947998, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8720393180847168, "num_tokens": 158349420.0, "step": 4146 }, { "epoch": 0.5275410253148455, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.029509544372559, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8567144870758057, "num_tokens": 158395834.0, "step": 4147 }, { "epoch": 0.5276682355934359, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.063674926757812, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.85890132188797, "num_tokens": 158435568.0, "step": 4148 }, { "epoch": 0.5277954458720264, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.070045471191406, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8672909736633301, "num_tokens": 158473584.0, "step": 4149 }, { "epoch": 0.527922656150617, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.005227088928223, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8585277795791626, "num_tokens": 158513435.0, "step": 4150 }, { "epoch": 0.5280498664292075, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.063868522644043, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.858799397945404, "num_tokens": 158549478.0, "step": 4151 }, { "epoch": 0.528177076707798, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 9.989673614501953, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8621333241462708, "num_tokens": 158598742.0, "step": 4152 }, { "epoch": 0.5283042869863885, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.058417320251465, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8466565608978271, "num_tokens": 158636806.0, "step": 4153 }, { "epoch": 0.528431497264979, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.08586597442627, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8516754508018494, "num_tokens": 158676043.0, "step": 4154 }, { "epoch": 0.5285587075435695, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.056031227111816, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8422173261642456, "num_tokens": 158711772.0, "step": 4155 }, { "epoch": 0.52868591782216, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.056365013122559, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8570713400840759, "num_tokens": 158749608.0, "step": 4156 }, { "epoch": 0.5288131281007505, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.041874885559082, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8590434789657593, "num_tokens": 158790502.0, "step": 4157 }, { "epoch": 0.5289403383793411, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.076324462890625, "learning_rate": 1e-06, "loss": 0.5154, "mean_token_accuracy": 0.8341221809387207, "num_tokens": 158828415.0, "step": 4158 }, { "epoch": 0.5290675486579316, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.111422538757324, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8484092950820923, "num_tokens": 158866972.0, "step": 4159 }, { "epoch": 0.5291947589365221, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.024330139160156, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8558661937713623, "num_tokens": 158902991.0, "step": 4160 }, { "epoch": 0.5293219692151125, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.049187660217285, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8654857873916626, "num_tokens": 158945963.0, "step": 4161 }, { "epoch": 0.5294491794937031, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.083467483520508, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8566884994506836, "num_tokens": 158985407.0, "step": 4162 }, { "epoch": 0.5295763897722936, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.069461822509766, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8507322072982788, "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": 10.0537109375, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8497915267944336, "num_tokens": 159061430.0, "step": 4164 }, { "epoch": 0.5298308103294747, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.051891326904297, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8597003221511841, "num_tokens": 159099350.0, "step": 4165 }, { "epoch": 0.5299580206080652, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.117694854736328, "learning_rate": 1e-06, "loss": 0.531, "mean_token_accuracy": 0.8328490257263184, "num_tokens": 159136276.0, "step": 4166 }, { "epoch": 0.5300852308866556, "ewc_loss": 0.017578125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 10.022379875183105, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8795458078384399, "num_tokens": 159177102.0, "step": 4167 }, { "epoch": 0.5302124411652461, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 10.079523086547852, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8697471618652344, "num_tokens": 159213582.0, "step": 4168 }, { "epoch": 0.5303396514438367, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.068909645080566, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8567061424255371, "num_tokens": 159254283.0, "step": 4169 }, { "epoch": 0.5304668617224272, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.062076568603516, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8597665429115295, "num_tokens": 159288557.0, "step": 4170 }, { "epoch": 0.5305940720010177, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.050979614257812, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8672835230827332, "num_tokens": 159325047.0, "step": 4171 }, { "epoch": 0.5307212822796082, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.040345191955566, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8617782592773438, "num_tokens": 159366291.0, "step": 4172 }, { "epoch": 0.5308484925581987, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.08222484588623, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8519856929779053, "num_tokens": 159401634.0, "step": 4173 }, { "epoch": 0.5309757028367892, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.106751441955566, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8493711948394775, "num_tokens": 159435316.0, "step": 4174 }, { "epoch": 0.5311029131153797, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.037148475646973, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.851190984249115, "num_tokens": 159472192.0, "step": 4175 }, { "epoch": 0.5312301233939702, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.079742431640625, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8648168444633484, "num_tokens": 159515342.0, "step": 4176 }, { "epoch": 0.5313573336725608, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.066988945007324, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8619919419288635, "num_tokens": 159553932.0, "step": 4177 }, { "epoch": 0.5314845439511513, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.083641052246094, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8594841957092285, "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": 10.098714828491211, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8448386788368225, "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": 10.074728965759277, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8562145233154297, "num_tokens": 159661288.0, "step": 4180 }, { "epoch": 0.5318661747869228, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.068650245666504, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8695752620697021, "num_tokens": 159701286.0, "step": 4181 }, { "epoch": 0.5319933850655133, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.052124977111816, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8459566831588745, "num_tokens": 159746631.0, "step": 4182 }, { "epoch": 0.5321205953441038, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.065511703491211, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8617977499961853, "num_tokens": 159787753.0, "step": 4183 }, { "epoch": 0.5322478056226944, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.047560691833496, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8501406908035278, "num_tokens": 159823092.0, "step": 4184 }, { "epoch": 0.5323750159012848, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.090435028076172, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8494988083839417, "num_tokens": 159858480.0, "step": 4185 }, { "epoch": 0.5325022261798753, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.09642505645752, "learning_rate": 1e-06, "loss": 0.4908, "mean_token_accuracy": 0.8482851982116699, "num_tokens": 159896220.0, "step": 4186 }, { "epoch": 0.5326294364584658, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.10009479522705, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.855033278465271, "num_tokens": 159930995.0, "step": 4187 }, { "epoch": 0.5327566467370564, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.103592872619629, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8531204462051392, "num_tokens": 159968013.0, "step": 4188 }, { "epoch": 0.5328838570156469, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.080907821655273, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8507245779037476, "num_tokens": 160004085.0, "step": 4189 }, { "epoch": 0.5330110672942374, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.13148021697998, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.868423581123352, "num_tokens": 160040606.0, "step": 4190 }, { "epoch": 0.5331382775728278, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.078181266784668, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8694106340408325, "num_tokens": 160080792.0, "step": 4191 }, { "epoch": 0.5332654878514184, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.070269584655762, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8460938930511475, "num_tokens": 160117055.0, "step": 4192 }, { "epoch": 0.5333926981300089, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.074995040893555, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8627496361732483, "num_tokens": 160153223.0, "step": 4193 }, { "epoch": 0.5335199084085994, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.034148216247559, "learning_rate": 1e-06, "loss": 0.4981, "mean_token_accuracy": 0.8404483795166016, "num_tokens": 160190920.0, "step": 4194 }, { "epoch": 0.53364711868719, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.137711524963379, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8550664186477661, "num_tokens": 160221898.0, "step": 4195 }, { "epoch": 0.5337743289657805, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.035235404968262, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.860143780708313, "num_tokens": 160267102.0, "step": 4196 }, { "epoch": 0.5339015392443709, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.104192733764648, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8629999756813049, "num_tokens": 160304422.0, "step": 4197 }, { "epoch": 0.5340287495229614, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.111666679382324, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8551703691482544, "num_tokens": 160339006.0, "step": 4198 }, { "epoch": 0.534155959801552, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.070569038391113, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8775515556335449, "num_tokens": 160377504.0, "step": 4199 }, { "epoch": 0.5342831700801425, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.087617874145508, "learning_rate": 1e-06, "loss": 0.4985, "mean_token_accuracy": 0.8374703526496887, "num_tokens": 160413315.0, "step": 4200 }, { "epoch": 0.534410380358733, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.038043022155762, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8716513514518738, "num_tokens": 160455034.0, "step": 4201 }, { "epoch": 0.5345375906373235, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.123059272766113, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8528358936309814, "num_tokens": 160490505.0, "step": 4202 }, { "epoch": 0.534664800915914, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.060028076171875, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8678776025772095, "num_tokens": 160528712.0, "step": 4203 }, { "epoch": 0.5347920111945045, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.085101127624512, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8499504923820496, "num_tokens": 160567945.0, "step": 4204 }, { "epoch": 0.534919221473095, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.105549812316895, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8647596836090088, "num_tokens": 160596767.0, "step": 4205 }, { "epoch": 0.5350464317516855, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.026952743530273, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8543888330459595, "num_tokens": 160636912.0, "step": 4206 }, { "epoch": 0.5351736420302761, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.100188255310059, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.851115345954895, "num_tokens": 160674588.0, "step": 4207 }, { "epoch": 0.5353008523088666, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.027572631835938, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8454564809799194, "num_tokens": 160713441.0, "step": 4208 }, { "epoch": 0.5354280625874571, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.112043380737305, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8677575588226318, "num_tokens": 160752716.0, "step": 4209 }, { "epoch": 0.5355552728660475, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.038576126098633, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8678480386734009, "num_tokens": 160798771.0, "step": 4210 }, { "epoch": 0.5356824831446381, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.082480430603027, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8751016855239868, "num_tokens": 160843085.0, "step": 4211 }, { "epoch": 0.5358096934232286, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.107423782348633, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8504688739776611, "num_tokens": 160883073.0, "step": 4212 }, { "epoch": 0.5359369037018191, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.080391883850098, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8647481203079224, "num_tokens": 160923734.0, "step": 4213 }, { "epoch": 0.5360641139804097, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.096131324768066, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8545613884925842, "num_tokens": 160963460.0, "step": 4214 }, { "epoch": 0.5361913242590002, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.130186080932617, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8553380966186523, "num_tokens": 161003691.0, "step": 4215 }, { "epoch": 0.5363185345375906, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.056251525878906, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.859314501285553, "num_tokens": 161042056.0, "step": 4216 }, { "epoch": 0.5364457448161811, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.116110801696777, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.881290853023529, "num_tokens": 161078670.0, "step": 4217 }, { "epoch": 0.5365729550947717, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.076868057250977, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8722434639930725, "num_tokens": 161113992.0, "step": 4218 }, { "epoch": 0.5367001653733622, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.081425666809082, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8637237548828125, "num_tokens": 161153763.0, "step": 4219 }, { "epoch": 0.5368273756519527, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.125438690185547, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8430139422416687, "num_tokens": 161184186.0, "step": 4220 }, { "epoch": 0.5369545859305432, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.104534149169922, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8561021089553833, "num_tokens": 161219298.0, "step": 4221 }, { "epoch": 0.5370817962091337, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.047194480895996, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8738483190536499, "num_tokens": 161255252.0, "step": 4222 }, { "epoch": 0.5372090064877242, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.153887748718262, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8613070845603943, "num_tokens": 161295890.0, "step": 4223 }, { "epoch": 0.5373362167663147, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.11366081237793, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8556621074676514, "num_tokens": 161330808.0, "step": 4224 }, { "epoch": 0.5374634270449052, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.076342582702637, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8664539456367493, "num_tokens": 161370515.0, "step": 4225 }, { "epoch": 0.5375906373234958, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.113139152526855, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8707929849624634, "num_tokens": 161405257.0, "step": 4226 }, { "epoch": 0.5377178476020863, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.068207740783691, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8590677976608276, "num_tokens": 161443526.0, "step": 4227 }, { "epoch": 0.5378450578806767, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.094802856445312, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8525204062461853, "num_tokens": 161481262.0, "step": 4228 }, { "epoch": 0.5379722681592672, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.140268325805664, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8435654640197754, "num_tokens": 161517756.0, "step": 4229 }, { "epoch": 0.5380994784378578, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.085888862609863, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8676956295967102, "num_tokens": 161552680.0, "step": 4230 }, { "epoch": 0.5382266887164483, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.135881423950195, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8555117845535278, "num_tokens": 161588607.0, "step": 4231 }, { "epoch": 0.5383538989950388, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.067008018493652, "learning_rate": 1e-06, "loss": 0.4944, "mean_token_accuracy": 0.8421416282653809, "num_tokens": 161626427.0, "step": 4232 }, { "epoch": 0.5384811092736294, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.099721908569336, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8642878532409668, "num_tokens": 161665182.0, "step": 4233 }, { "epoch": 0.5386083195522198, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.101860046386719, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8585203886032104, "num_tokens": 161704685.0, "step": 4234 }, { "epoch": 0.5387355298308103, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.08636474609375, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8638057708740234, "num_tokens": 161742960.0, "step": 4235 }, { "epoch": 0.5388627401094008, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.103267669677734, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8577908873558044, "num_tokens": 161778199.0, "step": 4236 }, { "epoch": 0.5389899503879914, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.13843822479248, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8542470932006836, "num_tokens": 161811235.0, "step": 4237 }, { "epoch": 0.5391171606665819, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.124785423278809, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8589688539505005, "num_tokens": 161847669.0, "step": 4238 }, { "epoch": 0.5392443709451724, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.09212589263916, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.8451166152954102, "num_tokens": 161888470.0, "step": 4239 }, { "epoch": 0.5393715812237628, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.174220085144043, "learning_rate": 1e-06, "loss": 0.4984, "mean_token_accuracy": 0.8396353721618652, "num_tokens": 161924563.0, "step": 4240 }, { "epoch": 0.5394987915023534, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.15814208984375, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.844282329082489, "num_tokens": 161959062.0, "step": 4241 }, { "epoch": 0.5396260017809439, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.073258399963379, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8626512289047241, "num_tokens": 161994755.0, "step": 4242 }, { "epoch": 0.5397532120595344, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.100415229797363, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8685476779937744, "num_tokens": 162032812.0, "step": 4243 }, { "epoch": 0.5398804223381249, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.071539878845215, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.861296534538269, "num_tokens": 162067130.0, "step": 4244 }, { "epoch": 0.5400076326167155, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.105730056762695, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8605554699897766, "num_tokens": 162107452.0, "step": 4245 }, { "epoch": 0.5401348428953059, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.154268264770508, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8581521511077881, "num_tokens": 162144220.0, "step": 4246 }, { "epoch": 0.5402620531738964, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.086088180541992, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8681602478027344, "num_tokens": 162181798.0, "step": 4247 }, { "epoch": 0.540389263452487, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.070611953735352, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8508148193359375, "num_tokens": 162222501.0, "step": 4248 }, { "epoch": 0.5405164737310775, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.182892799377441, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8551781177520752, "num_tokens": 162265533.0, "step": 4249 }, { "epoch": 0.540643684009668, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.087241172790527, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8736659288406372, "num_tokens": 162304984.0, "step": 4250 }, { "epoch": 0.5407708942882585, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.12516975402832, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.849999725818634, "num_tokens": 162342358.0, "step": 4251 }, { "epoch": 0.540898104566849, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.106585502624512, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8600724935531616, "num_tokens": 162377343.0, "step": 4252 }, { "epoch": 0.5410253148454395, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.05138874053955, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8596967458724976, "num_tokens": 162416416.0, "step": 4253 }, { "epoch": 0.54115252512403, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.178744316101074, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8843882083892822, "num_tokens": 162447108.0, "step": 4254 }, { "epoch": 0.5412797354026205, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.080038070678711, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8603810667991638, "num_tokens": 162489659.0, "step": 4255 }, { "epoch": 0.5414069456812111, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.156353950500488, "learning_rate": 1e-06, "loss": 0.4719, "mean_token_accuracy": 0.8493905067443848, "num_tokens": 162525816.0, "step": 4256 }, { "epoch": 0.5415341559598016, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.109025955200195, "learning_rate": 1e-06, "loss": 0.4757, "mean_token_accuracy": 0.8548827171325684, "num_tokens": 162564154.0, "step": 4257 }, { "epoch": 0.5416613662383921, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.121638298034668, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8524607419967651, "num_tokens": 162602728.0, "step": 4258 }, { "epoch": 0.5417885765169825, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.112810134887695, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8636839985847473, "num_tokens": 162636659.0, "step": 4259 }, { "epoch": 0.5419157867955731, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.135393142700195, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.87513267993927, "num_tokens": 162676639.0, "step": 4260 }, { "epoch": 0.5420429970741636, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.090592384338379, "learning_rate": 1e-06, "loss": 0.4966, "mean_token_accuracy": 0.8403761982917786, "num_tokens": 162716763.0, "step": 4261 }, { "epoch": 0.5421702073527541, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.19001579284668, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8495159149169922, "num_tokens": 162753943.0, "step": 4262 }, { "epoch": 0.5422974176313446, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.11443042755127, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.853823184967041, "num_tokens": 162792938.0, "step": 4263 }, { "epoch": 0.5424246279099352, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 10.07597541809082, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8563973903656006, "num_tokens": 162838938.0, "step": 4264 }, { "epoch": 0.5425518381885256, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.130644798278809, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8646420836448669, "num_tokens": 162875026.0, "step": 4265 }, { "epoch": 0.5426790484671161, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.11131477355957, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8619738817214966, "num_tokens": 162910724.0, "step": 4266 }, { "epoch": 0.5428062587457066, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.146533012390137, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8617604374885559, "num_tokens": 162941602.0, "step": 4267 }, { "epoch": 0.5429334690242972, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.142817497253418, "learning_rate": 1e-06, "loss": 0.4877, "mean_token_accuracy": 0.8457373380661011, "num_tokens": 162980760.0, "step": 4268 }, { "epoch": 0.5430606793028877, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.090279579162598, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8691266775131226, "num_tokens": 163016155.0, "step": 4269 }, { "epoch": 0.5431878895814782, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.170014381408691, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8505508899688721, "num_tokens": 163052519.0, "step": 4270 }, { "epoch": 0.5433150998600687, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.156281471252441, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8673888444900513, "num_tokens": 163091806.0, "step": 4271 }, { "epoch": 0.5434423101386592, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.139128684997559, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8745851516723633, "num_tokens": 163128830.0, "step": 4272 }, { "epoch": 0.5435695204172497, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.046394348144531, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8539657592773438, "num_tokens": 163169567.0, "step": 4273 }, { "epoch": 0.5436967306958402, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.12358283996582, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8521589040756226, "num_tokens": 163202913.0, "step": 4274 }, { "epoch": 0.5438239409744308, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.100628852844238, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8668268918991089, "num_tokens": 163236741.0, "step": 4275 }, { "epoch": 0.5439511512530213, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.141731262207031, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8527532815933228, "num_tokens": 163281402.0, "step": 4276 }, { "epoch": 0.5440783615316117, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.061614990234375, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8678315281867981, "num_tokens": 163325288.0, "step": 4277 }, { "epoch": 0.5442055718102022, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.112268447875977, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8485682010650635, "num_tokens": 163367471.0, "step": 4278 }, { "epoch": 0.5443327820887928, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.137884140014648, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8608015775680542, "num_tokens": 163400077.0, "step": 4279 }, { "epoch": 0.5444599923673833, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.11876106262207, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8614970445632935, "num_tokens": 163433739.0, "step": 4280 }, { "epoch": 0.5445872026459738, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.12906551361084, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8598743677139282, "num_tokens": 163467447.0, "step": 4281 }, { "epoch": 0.5447144129245644, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.108644485473633, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8560107350349426, "num_tokens": 163506633.0, "step": 4282 }, { "epoch": 0.5448416232031548, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.1581449508667, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8435055017471313, "num_tokens": 163542773.0, "step": 4283 }, { "epoch": 0.5449688334817453, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.137187004089355, "learning_rate": 1e-06, "loss": 0.5415, "mean_token_accuracy": 0.8249785900115967, "num_tokens": 163585247.0, "step": 4284 }, { "epoch": 0.5450960437603358, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.094404220581055, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8469615578651428, "num_tokens": 163622802.0, "step": 4285 }, { "epoch": 0.5452232540389264, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.144770622253418, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8830088376998901, "num_tokens": 163655980.0, "step": 4286 }, { "epoch": 0.5453504643175169, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.088372230529785, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8641554117202759, "num_tokens": 163691494.0, "step": 4287 }, { "epoch": 0.5454776745961074, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.189729690551758, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8701661229133606, "num_tokens": 163725169.0, "step": 4288 }, { "epoch": 0.5456048848746978, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.117505073547363, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8636529445648193, "num_tokens": 163766050.0, "step": 4289 }, { "epoch": 0.5457320951532884, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.156976699829102, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8535844087600708, "num_tokens": 163799559.0, "step": 4290 }, { "epoch": 0.5458593054318789, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.126555442810059, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8533060550689697, "num_tokens": 163837052.0, "step": 4291 }, { "epoch": 0.5459865157104694, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.171133995056152, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8564291000366211, "num_tokens": 163872494.0, "step": 4292 }, { "epoch": 0.5461137259890599, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.103006362915039, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8510119915008545, "num_tokens": 163913499.0, "step": 4293 }, { "epoch": 0.5462409362676505, "ewc_loss": 0.017822265625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 10.127954483032227, "learning_rate": 1e-06, "loss": 0.4967, "mean_token_accuracy": 0.8396925926208496, "num_tokens": 163957965.0, "step": 4294 }, { "epoch": 0.5463681465462409, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.133634567260742, "learning_rate": 1e-06, "loss": 0.5688, "mean_token_accuracy": 0.8185364603996277, "num_tokens": 164005661.0, "step": 4295 }, { "epoch": 0.5464953568248314, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.13012409210205, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8743213415145874, "num_tokens": 164044717.0, "step": 4296 }, { "epoch": 0.5466225671034219, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.143620491027832, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8588762283325195, "num_tokens": 164080022.0, "step": 4297 }, { "epoch": 0.5467497773820125, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.17010498046875, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8651318550109863, "num_tokens": 164112749.0, "step": 4298 }, { "epoch": 0.546876987660603, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.125419616699219, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8531752824783325, "num_tokens": 164152517.0, "step": 4299 }, { "epoch": 0.5470041979391935, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.11556625366211, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8653340935707092, "num_tokens": 164195245.0, "step": 4300 }, { "epoch": 0.5471314082177839, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.150894165039062, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8541151285171509, "num_tokens": 164230643.0, "step": 4301 }, { "epoch": 0.5472586184963745, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.09406852722168, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8550140857696533, "num_tokens": 164271089.0, "step": 4302 }, { "epoch": 0.547385828774965, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.137751579284668, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8569119572639465, "num_tokens": 164310618.0, "step": 4303 }, { "epoch": 0.5475130390535555, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.154294967651367, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8449345231056213, "num_tokens": 164348630.0, "step": 4304 }, { "epoch": 0.5476402493321461, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.087117195129395, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8650850057601929, "num_tokens": 164383735.0, "step": 4305 }, { "epoch": 0.5477674596107366, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.111187934875488, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8601089715957642, "num_tokens": 164426115.0, "step": 4306 }, { "epoch": 0.5478946698893271, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.144207000732422, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8514788150787354, "num_tokens": 164462119.0, "step": 4307 }, { "epoch": 0.5480218801679175, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.10896110534668, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8554156422615051, "num_tokens": 164503030.0, "step": 4308 }, { "epoch": 0.5481490904465081, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.113170623779297, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8615492582321167, "num_tokens": 164541091.0, "step": 4309 }, { "epoch": 0.5482763007250986, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.126175880432129, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.853289008140564, "num_tokens": 164577678.0, "step": 4310 }, { "epoch": 0.5484035110036891, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.145021438598633, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8551628589630127, "num_tokens": 164612779.0, "step": 4311 }, { "epoch": 0.5485307212822796, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.105138778686523, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8633381724357605, "num_tokens": 164649849.0, "step": 4312 }, { "epoch": 0.5486579315608702, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.124834060668945, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8554288148880005, "num_tokens": 164690538.0, "step": 4313 }, { "epoch": 0.5487851418394606, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.161797523498535, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8594173192977905, "num_tokens": 164729338.0, "step": 4314 }, { "epoch": 0.5489123521180511, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.13891315460205, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8701570630073547, "num_tokens": 164766648.0, "step": 4315 }, { "epoch": 0.5490395623966416, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.140902519226074, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8578478097915649, "num_tokens": 164804254.0, "step": 4316 }, { "epoch": 0.5491667726752322, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.090746879577637, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8710008859634399, "num_tokens": 164839943.0, "step": 4317 }, { "epoch": 0.5492939829538227, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.090853691101074, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8510477542877197, "num_tokens": 164887057.0, "step": 4318 }, { "epoch": 0.5494211932324132, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.144416809082031, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8572174906730652, "num_tokens": 164928515.0, "step": 4319 }, { "epoch": 0.5495484035110036, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.143909454345703, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8579258918762207, "num_tokens": 164963818.0, "step": 4320 }, { "epoch": 0.5496756137895942, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.152620315551758, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8724284768104553, "num_tokens": 165001091.0, "step": 4321 }, { "epoch": 0.5498028240681847, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.100719451904297, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8413457870483398, "num_tokens": 165046009.0, "step": 4322 }, { "epoch": 0.5499300343467752, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.136557579040527, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8549556732177734, "num_tokens": 165090729.0, "step": 4323 }, { "epoch": 0.5500572446253658, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.097058296203613, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8755695819854736, "num_tokens": 165129771.0, "step": 4324 }, { "epoch": 0.5501844549039563, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.150235176086426, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.874355137348175, "num_tokens": 165164310.0, "step": 4325 }, { "epoch": 0.5503116651825467, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.163865089416504, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8556264638900757, "num_tokens": 165202842.0, "step": 4326 }, { "epoch": 0.5504388754611372, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.152497291564941, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8621616959571838, "num_tokens": 165235998.0, "step": 4327 }, { "epoch": 0.5505660857397278, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.177501678466797, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8590995669364929, "num_tokens": 165271779.0, "step": 4328 }, { "epoch": 0.5506932960183183, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.127528190612793, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8528354167938232, "num_tokens": 165306136.0, "step": 4329 }, { "epoch": 0.5508205062969088, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.199298858642578, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.872887134552002, "num_tokens": 165342401.0, "step": 4330 }, { "epoch": 0.5509477165754993, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.155077934265137, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8619563579559326, "num_tokens": 165380120.0, "step": 4331 }, { "epoch": 0.5510749268540898, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.165508270263672, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8461934924125671, "num_tokens": 165418210.0, "step": 4332 }, { "epoch": 0.5512021371326803, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.201255798339844, "learning_rate": 1e-06, "loss": 0.5104, "mean_token_accuracy": 0.8380957841873169, "num_tokens": 165460952.0, "step": 4333 }, { "epoch": 0.5513293474112708, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.13170337677002, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8715389370918274, "num_tokens": 165500370.0, "step": 4334 }, { "epoch": 0.5514565576898613, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.155855178833008, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.859451949596405, "num_tokens": 165534089.0, "step": 4335 }, { "epoch": 0.5515837679684519, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.163792610168457, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.851449728012085, "num_tokens": 165573111.0, "step": 4336 }, { "epoch": 0.5517109782470424, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.135486602783203, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8492776155471802, "num_tokens": 165611942.0, "step": 4337 }, { "epoch": 0.5518381885256328, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.151084899902344, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8659733533859253, "num_tokens": 165651703.0, "step": 4338 }, { "epoch": 0.5519653988042234, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.145323753356934, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8566690683364868, "num_tokens": 165691257.0, "step": 4339 }, { "epoch": 0.5520926090828139, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 10.183523178100586, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8742055892944336, "num_tokens": 165727947.0, "step": 4340 }, { "epoch": 0.5522198193614044, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.119733810424805, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8677572011947632, "num_tokens": 165763400.0, "step": 4341 }, { "epoch": 0.5523470296399949, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.17430591583252, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8590199947357178, "num_tokens": 165809586.0, "step": 4342 }, { "epoch": 0.5524742399185855, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.157033920288086, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8612247705459595, "num_tokens": 165839565.0, "step": 4343 }, { "epoch": 0.5526014501971759, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.197183609008789, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8624116778373718, "num_tokens": 165873539.0, "step": 4344 }, { "epoch": 0.5527286604757664, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.0755033493042, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8676822781562805, "num_tokens": 165920987.0, "step": 4345 }, { "epoch": 0.5528558707543569, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.150999069213867, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8486219644546509, "num_tokens": 165965661.0, "step": 4346 }, { "epoch": 0.5529830810329475, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.166699409484863, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8523406982421875, "num_tokens": 165998239.0, "step": 4347 }, { "epoch": 0.553110291311538, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.10859489440918, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.85645592212677, "num_tokens": 166036489.0, "step": 4348 }, { "epoch": 0.5532375015901285, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.140220642089844, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8554213047027588, "num_tokens": 166078774.0, "step": 4349 }, { "epoch": 0.5533647118687189, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.118337631225586, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8652585744857788, "num_tokens": 166119555.0, "step": 4350 }, { "epoch": 0.5534919221473095, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.14893627166748, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8570642471313477, "num_tokens": 166159788.0, "step": 4351 }, { "epoch": 0.5536191324259, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.131590843200684, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8533326387405396, "num_tokens": 166200230.0, "step": 4352 }, { "epoch": 0.5537463427044905, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.13155746459961, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8620626926422119, "num_tokens": 166237864.0, "step": 4353 }, { "epoch": 0.553873552983081, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.17458724975586, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8617519736289978, "num_tokens": 166278347.0, "step": 4354 }, { "epoch": 0.5540007632616716, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.153411865234375, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8499365448951721, "num_tokens": 166315744.0, "step": 4355 }, { "epoch": 0.554127973540262, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.17892074584961, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8602749109268188, "num_tokens": 166352705.0, "step": 4356 }, { "epoch": 0.5542551838188525, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.124011039733887, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8635467290878296, "num_tokens": 166393956.0, "step": 4357 }, { "epoch": 0.554382394097443, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.21440601348877, "learning_rate": 1e-06, "loss": 0.5206, "mean_token_accuracy": 0.8358231782913208, "num_tokens": 166430172.0, "step": 4358 }, { "epoch": 0.5545096043760336, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.16353988647461, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.859040379524231, "num_tokens": 166468732.0, "step": 4359 }, { "epoch": 0.5546368146546241, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.123387336730957, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8548286557197571, "num_tokens": 166511383.0, "step": 4360 }, { "epoch": 0.5547640249332146, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.155716896057129, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8631299734115601, "num_tokens": 166547674.0, "step": 4361 }, { "epoch": 0.5548912352118052, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.18072509765625, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.852040708065033, "num_tokens": 166581763.0, "step": 4362 }, { "epoch": 0.5550184454903956, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.151521682739258, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8560906052589417, "num_tokens": 166618604.0, "step": 4363 }, { "epoch": 0.5551456557689861, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.135589599609375, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8683087825775146, "num_tokens": 166654812.0, "step": 4364 }, { "epoch": 0.5552728660475766, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.170567512512207, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8449442982673645, "num_tokens": 166693206.0, "step": 4365 }, { "epoch": 0.5554000763261672, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.171704292297363, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8578044772148132, "num_tokens": 166734531.0, "step": 4366 }, { "epoch": 0.5555272866047577, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.150604248046875, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8567761182785034, "num_tokens": 166771362.0, "step": 4367 }, { "epoch": 0.5556544968833482, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.202740669250488, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.849183201789856, "num_tokens": 166802784.0, "step": 4368 }, { "epoch": 0.5557817071619386, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.129927635192871, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8539254665374756, "num_tokens": 166835590.0, "step": 4369 }, { "epoch": 0.5559089174405292, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.150120735168457, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.853904128074646, "num_tokens": 166872145.0, "step": 4370 }, { "epoch": 0.5560361277191197, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.13183879852295, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8552339673042297, "num_tokens": 166912855.0, "step": 4371 }, { "epoch": 0.5561633379977102, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.16846752166748, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8432323932647705, "num_tokens": 166949482.0, "step": 4372 }, { "epoch": 0.5562905482763008, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.187606811523438, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8585933446884155, "num_tokens": 166986727.0, "step": 4373 }, { "epoch": 0.5564177585548913, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.170217514038086, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8516008853912354, "num_tokens": 167025869.0, "step": 4374 }, { "epoch": 0.5565449688334817, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.169998168945312, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8687087893486023, "num_tokens": 167059825.0, "step": 4375 }, { "epoch": 0.5566721791120722, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.177589416503906, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.862438440322876, "num_tokens": 167097375.0, "step": 4376 }, { "epoch": 0.5567993893906628, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.214249610900879, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8495457768440247, "num_tokens": 167136555.0, "step": 4377 }, { "epoch": 0.5569265996692533, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.16675090789795, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8566309213638306, "num_tokens": 167179870.0, "step": 4378 }, { "epoch": 0.5570538099478438, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.159416198730469, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8572360277175903, "num_tokens": 167215242.0, "step": 4379 }, { "epoch": 0.5571810202264343, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.157279014587402, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8655410408973694, "num_tokens": 167250690.0, "step": 4380 }, { "epoch": 0.5573082305050248, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.221006393432617, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8465146422386169, "num_tokens": 167283677.0, "step": 4381 }, { "epoch": 0.5574354407836153, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.140887260437012, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8598122596740723, "num_tokens": 167321355.0, "step": 4382 }, { "epoch": 0.5575626510622058, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.141650199890137, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8501325845718384, "num_tokens": 167363103.0, "step": 4383 }, { "epoch": 0.5576898613407963, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.182110786437988, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8719315528869629, "num_tokens": 167399732.0, "step": 4384 }, { "epoch": 0.5578170716193869, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.164046287536621, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8566139936447144, "num_tokens": 167434938.0, "step": 4385 }, { "epoch": 0.5579442818979774, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.198968887329102, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8612028956413269, "num_tokens": 167472293.0, "step": 4386 }, { "epoch": 0.5580714921765678, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.135323524475098, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8629412055015564, "num_tokens": 167511015.0, "step": 4387 }, { "epoch": 0.5581987024551583, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.162944793701172, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8576167821884155, "num_tokens": 167549777.0, "step": 4388 }, { "epoch": 0.5583259127337489, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.178142547607422, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8647654056549072, "num_tokens": 167592342.0, "step": 4389 }, { "epoch": 0.5584531230123394, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.147547721862793, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8630001544952393, "num_tokens": 167630829.0, "step": 4390 }, { "epoch": 0.5585803332909299, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.171378135681152, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8468530774116516, "num_tokens": 167675938.0, "step": 4391 }, { "epoch": 0.5587075435695205, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.161439895629883, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8626284599304199, "num_tokens": 167716550.0, "step": 4392 }, { "epoch": 0.5588347538481109, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.17774772644043, "learning_rate": 1e-06, "loss": 0.4846, "mean_token_accuracy": 0.8453347086906433, "num_tokens": 167754401.0, "step": 4393 }, { "epoch": 0.5589619641267014, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.180929183959961, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8622618317604065, "num_tokens": 167792124.0, "step": 4394 }, { "epoch": 0.5590891744052919, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.15078067779541, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8762930035591125, "num_tokens": 167829226.0, "step": 4395 }, { "epoch": 0.5592163846838825, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.17815113067627, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8690711259841919, "num_tokens": 167863439.0, "step": 4396 }, { "epoch": 0.559343594962473, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.128344535827637, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.858595609664917, "num_tokens": 167907019.0, "step": 4397 }, { "epoch": 0.5594708052410635, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.202130317687988, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8505174517631531, "num_tokens": 167942539.0, "step": 4398 }, { "epoch": 0.5595980155196539, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.133797645568848, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8583433628082275, "num_tokens": 167982848.0, "step": 4399 }, { "epoch": 0.5597252257982445, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.192651748657227, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8677071332931519, "num_tokens": 168017500.0, "step": 4400 }, { "epoch": 0.559852436076835, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.212834358215332, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8438056707382202, "num_tokens": 168052790.0, "step": 4401 }, { "epoch": 0.5599796463554255, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.164891242980957, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8659457564353943, "num_tokens": 168082541.0, "step": 4402 }, { "epoch": 0.560106856634016, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.160463333129883, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8645299673080444, "num_tokens": 168128218.0, "step": 4403 }, { "epoch": 0.5602340669126066, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.209310531616211, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8613322377204895, "num_tokens": 168165496.0, "step": 4404 }, { "epoch": 0.560361277191197, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.173696517944336, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8455508947372437, "num_tokens": 168204675.0, "step": 4405 }, { "epoch": 0.5604884874697875, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.206533432006836, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8547991514205933, "num_tokens": 168244832.0, "step": 4406 }, { "epoch": 0.560615697748378, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.183257102966309, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8544724583625793, "num_tokens": 168282440.0, "step": 4407 }, { "epoch": 0.5607429080269686, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.165743827819824, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8640981912612915, "num_tokens": 168325303.0, "step": 4408 }, { "epoch": 0.5608701183055591, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.22867488861084, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8645904064178467, "num_tokens": 168360628.0, "step": 4409 }, { "epoch": 0.5609973285841496, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.153257369995117, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8461385369300842, "num_tokens": 168398601.0, "step": 4410 }, { "epoch": 0.5611245388627402, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.171338081359863, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.844978928565979, "num_tokens": 168436893.0, "step": 4411 }, { "epoch": 0.5612517491413306, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.134208679199219, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8542402386665344, "num_tokens": 168478794.0, "step": 4412 }, { "epoch": 0.5613789594199211, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.210367202758789, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8559066653251648, "num_tokens": 168520184.0, "step": 4413 }, { "epoch": 0.5615061696985116, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.169239044189453, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8550703525543213, "num_tokens": 168554305.0, "step": 4414 }, { "epoch": 0.5616333799771022, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.151457786560059, "learning_rate": 1e-06, "loss": 0.4935, "mean_token_accuracy": 0.842636227607727, "num_tokens": 168597752.0, "step": 4415 }, { "epoch": 0.5617605902556927, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.220019340515137, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8642925024032593, "num_tokens": 168637764.0, "step": 4416 }, { "epoch": 0.5618878005342832, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.228121757507324, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8389800786972046, "num_tokens": 168672260.0, "step": 4417 }, { "epoch": 0.5620150108128736, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.173028945922852, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8666940927505493, "num_tokens": 168711478.0, "step": 4418 }, { "epoch": 0.5621422210914642, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.175155639648438, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8613893985748291, "num_tokens": 168751773.0, "step": 4419 }, { "epoch": 0.5622694313700547, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.204014778137207, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8688526153564453, "num_tokens": 168786364.0, "step": 4420 }, { "epoch": 0.5623966416486452, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.199602127075195, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8552373051643372, "num_tokens": 168826819.0, "step": 4421 }, { "epoch": 0.5625238519272358, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.170530319213867, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8562164902687073, "num_tokens": 168869405.0, "step": 4422 }, { "epoch": 0.5626510622058263, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.213933944702148, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8647467494010925, "num_tokens": 168904323.0, "step": 4423 }, { "epoch": 0.5627782724844167, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.18452262878418, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8490350246429443, "num_tokens": 168948654.0, "step": 4424 }, { "epoch": 0.5629054827630072, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.195062637329102, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8647312521934509, "num_tokens": 168984333.0, "step": 4425 }, { "epoch": 0.5630326930415978, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.221160888671875, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8592213988304138, "num_tokens": 169019843.0, "step": 4426 }, { "epoch": 0.5631599033201883, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.215835571289062, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8692647814750671, "num_tokens": 169058177.0, "step": 4427 }, { "epoch": 0.5632871135987788, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.227882385253906, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8645905256271362, "num_tokens": 169102268.0, "step": 4428 }, { "epoch": 0.5634143238773693, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.18292236328125, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8534802198410034, "num_tokens": 169141709.0, "step": 4429 }, { "epoch": 0.5635415341559598, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.26632022857666, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8552513122558594, "num_tokens": 169177107.0, "step": 4430 }, { "epoch": 0.5636687444345503, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.22661018371582, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8551071882247925, "num_tokens": 169211531.0, "step": 4431 }, { "epoch": 0.5637959547131408, "ewc_loss": 0.01806640625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 10.153877258300781, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8505042791366577, "num_tokens": 169247507.0, "step": 4432 }, { "epoch": 0.5639231649917313, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.261722564697266, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8723798990249634, "num_tokens": 169281007.0, "step": 4433 }, { "epoch": 0.5640503752703219, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.200278282165527, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8591750860214233, "num_tokens": 169322295.0, "step": 4434 }, { "epoch": 0.5641775855489124, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.214332580566406, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8552923202514648, "num_tokens": 169358184.0, "step": 4435 }, { "epoch": 0.5643047958275028, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.249990463256836, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8638700842857361, "num_tokens": 169391518.0, "step": 4436 }, { "epoch": 0.5644320061060933, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.180747985839844, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8586046695709229, "num_tokens": 169426743.0, "step": 4437 }, { "epoch": 0.5645592163846839, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.165922164916992, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8777631521224976, "num_tokens": 169466452.0, "step": 4438 }, { "epoch": 0.5646864266632744, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.229498863220215, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8634807467460632, "num_tokens": 169500295.0, "step": 4439 }, { "epoch": 0.5648136369418649, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.187430381774902, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8583548069000244, "num_tokens": 169541829.0, "step": 4440 }, { "epoch": 0.5649408472204555, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.195903778076172, "learning_rate": 1e-06, "loss": 0.4888, "mean_token_accuracy": 0.8430799841880798, "num_tokens": 169581547.0, "step": 4441 }, { "epoch": 0.5650680574990459, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.242851257324219, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8631959557533264, "num_tokens": 169622107.0, "step": 4442 }, { "epoch": 0.5651952677776364, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.172085762023926, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8560283184051514, "num_tokens": 169657258.0, "step": 4443 }, { "epoch": 0.5653224780562269, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.172435760498047, "learning_rate": 1e-06, "loss": 0.4801, "mean_token_accuracy": 0.8482544422149658, "num_tokens": 169691568.0, "step": 4444 }, { "epoch": 0.5654496883348175, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.233226776123047, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8648401498794556, "num_tokens": 169732429.0, "step": 4445 }, { "epoch": 0.565576898613408, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.186419486999512, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8514482975006104, "num_tokens": 169769224.0, "step": 4446 }, { "epoch": 0.5657041088919985, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.127071380615234, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8586994409561157, "num_tokens": 169812830.0, "step": 4447 }, { "epoch": 0.5658313191705889, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.271015167236328, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8558470010757446, "num_tokens": 169843823.0, "step": 4448 }, { "epoch": 0.5659585294491795, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.170466423034668, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8602445125579834, "num_tokens": 169882847.0, "step": 4449 }, { "epoch": 0.56608573972777, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.196284294128418, "learning_rate": 1e-06, "loss": 0.493, "mean_token_accuracy": 0.840090811252594, "num_tokens": 169921264.0, "step": 4450 }, { "epoch": 0.5662129500063605, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.231627464294434, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8650404214859009, "num_tokens": 169949681.0, "step": 4451 }, { "epoch": 0.566340160284951, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.214354515075684, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8725904226303101, "num_tokens": 169986254.0, "step": 4452 }, { "epoch": 0.5664673705635416, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.158743858337402, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8551720380783081, "num_tokens": 170022745.0, "step": 4453 }, { "epoch": 0.566594580842132, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.202707290649414, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8691365718841553, "num_tokens": 170057038.0, "step": 4454 }, { "epoch": 0.5667217911207225, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.188026428222656, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8776047825813293, "num_tokens": 170098460.0, "step": 4455 }, { "epoch": 0.566849001399313, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.169665336608887, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8458961248397827, "num_tokens": 170137689.0, "step": 4456 }, { "epoch": 0.5669762116779036, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.199189186096191, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8592884540557861, "num_tokens": 170176716.0, "step": 4457 }, { "epoch": 0.5671034219564941, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.279520988464355, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8544694185256958, "num_tokens": 170212621.0, "step": 4458 }, { "epoch": 0.5672306322350846, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.154938697814941, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8486027121543884, "num_tokens": 170254493.0, "step": 4459 }, { "epoch": 0.5673578425136752, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.238685607910156, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8803507089614868, "num_tokens": 170290559.0, "step": 4460 }, { "epoch": 0.5674850527922656, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.197725296020508, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.852533221244812, "num_tokens": 170335079.0, "step": 4461 }, { "epoch": 0.5676122630708561, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.242581367492676, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.861924409866333, "num_tokens": 170370665.0, "step": 4462 }, { "epoch": 0.5677394733494466, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.239102363586426, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8575308918952942, "num_tokens": 170412585.0, "step": 4463 }, { "epoch": 0.5678666836280372, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.243767738342285, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8649271726608276, "num_tokens": 170450806.0, "step": 4464 }, { "epoch": 0.5679938939066277, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.288810729980469, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8589548468589783, "num_tokens": 170483265.0, "step": 4465 }, { "epoch": 0.5681211041852182, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.272134780883789, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8755958080291748, "num_tokens": 170520910.0, "step": 4466 }, { "epoch": 0.5682483144638086, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.254942893981934, "learning_rate": 1e-06, "loss": 0.4909, "mean_token_accuracy": 0.8420016169548035, "num_tokens": 170556516.0, "step": 4467 }, { "epoch": 0.5683755247423992, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.277998924255371, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8563925623893738, "num_tokens": 170595571.0, "step": 4468 }, { "epoch": 0.5685027350209897, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.230710983276367, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8780015110969543, "num_tokens": 170634117.0, "step": 4469 }, { "epoch": 0.5686299452995802, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.24290657043457, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8539106845855713, "num_tokens": 170677624.0, "step": 4470 }, { "epoch": 0.5687571555781707, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.268315315246582, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8446274995803833, "num_tokens": 170716732.0, "step": 4471 }, { "epoch": 0.5688843658567613, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.277569770812988, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8527930378913879, "num_tokens": 170758521.0, "step": 4472 }, { "epoch": 0.5690115761353517, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.291534423828125, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8575655221939087, "num_tokens": 170796799.0, "step": 4473 }, { "epoch": 0.5691387864139422, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.296009063720703, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.845390796661377, "num_tokens": 170838569.0, "step": 4474 }, { "epoch": 0.5692659966925327, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.259567260742188, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8598797917366028, "num_tokens": 170878206.0, "step": 4475 }, { "epoch": 0.5693932069711233, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.310151100158691, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.8461676836013794, "num_tokens": 170915506.0, "step": 4476 }, { "epoch": 0.5695204172497138, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.233772277832031, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8547582626342773, "num_tokens": 170952303.0, "step": 4477 }, { "epoch": 0.5696476275283043, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.219108581542969, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8522607684135437, "num_tokens": 170993757.0, "step": 4478 }, { "epoch": 0.5697748378068948, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.2553129196167, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.856020450592041, "num_tokens": 171030240.0, "step": 4479 }, { "epoch": 0.5699020480854853, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.277978897094727, "learning_rate": 1e-06, "loss": 0.5029, "mean_token_accuracy": 0.8358602523803711, "num_tokens": 171069654.0, "step": 4480 }, { "epoch": 0.5700292583640758, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.216938018798828, "learning_rate": 1e-06, "loss": 0.491, "mean_token_accuracy": 0.8419899940490723, "num_tokens": 171107341.0, "step": 4481 }, { "epoch": 0.5701564686426663, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.239986419677734, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.849804699420929, "num_tokens": 171144063.0, "step": 4482 }, { "epoch": 0.5702836789212569, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.204798698425293, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.852733850479126, "num_tokens": 171180448.0, "step": 4483 }, { "epoch": 0.5704108891998474, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.299459457397461, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8628696203231812, "num_tokens": 171214220.0, "step": 4484 }, { "epoch": 0.5705380994784378, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.225077629089355, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8644734621047974, "num_tokens": 171252798.0, "step": 4485 }, { "epoch": 0.5706653097570283, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.251290321350098, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8633768558502197, "num_tokens": 171295434.0, "step": 4486 }, { "epoch": 0.5707925200356189, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.217498779296875, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.859687328338623, "num_tokens": 171334129.0, "step": 4487 }, { "epoch": 0.5709197303142094, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.279967308044434, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.854444146156311, "num_tokens": 171370239.0, "step": 4488 }, { "epoch": 0.5710469405927999, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.227241516113281, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8729133009910583, "num_tokens": 171402411.0, "step": 4489 }, { "epoch": 0.5711741508713905, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.258838653564453, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8684748411178589, "num_tokens": 171438736.0, "step": 4490 }, { "epoch": 0.5713013611499809, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.234455108642578, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8647854328155518, "num_tokens": 171477949.0, "step": 4491 }, { "epoch": 0.5714285714285714, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.25252628326416, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8539658188819885, "num_tokens": 171520953.0, "step": 4492 }, { "epoch": 0.5715557817071619, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.297689437866211, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8645539879798889, "num_tokens": 171559415.0, "step": 4493 }, { "epoch": 0.5716829919857525, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.252340316772461, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8659160137176514, "num_tokens": 171593118.0, "step": 4494 }, { "epoch": 0.571810202264343, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.25051212310791, "learning_rate": 1e-06, "loss": 0.4857, "mean_token_accuracy": 0.8454470634460449, "num_tokens": 171634514.0, "step": 4495 }, { "epoch": 0.5719374125429335, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.230125427246094, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8616121411323547, "num_tokens": 171675239.0, "step": 4496 }, { "epoch": 0.5720646228215239, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 10.231436729431152, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8664383888244629, "num_tokens": 171713154.0, "step": 4497 }, { "epoch": 0.5721918331001145, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.27115249633789, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8515778183937073, "num_tokens": 171750046.0, "step": 4498 }, { "epoch": 0.572319043378705, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.230117797851562, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8485676646232605, "num_tokens": 171782708.0, "step": 4499 }, { "epoch": 0.5724462536572955, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.248834609985352, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8548392057418823, "num_tokens": 171819421.0, "step": 4500 }, { "epoch": 0.572573463935886, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.226566314697266, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8577116131782532, "num_tokens": 171860993.0, "step": 4501 }, { "epoch": 0.5727006742144766, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.220001220703125, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8578488230705261, "num_tokens": 171896894.0, "step": 4502 }, { "epoch": 0.572827884493067, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.246428489685059, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8559554815292358, "num_tokens": 171938694.0, "step": 4503 }, { "epoch": 0.5729550947716575, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.255073547363281, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8830287456512451, "num_tokens": 171972606.0, "step": 4504 }, { "epoch": 0.573082305050248, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.259743690490723, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8606207370758057, "num_tokens": 172010861.0, "step": 4505 }, { "epoch": 0.5732095153288386, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.28217601776123, "learning_rate": 1e-06, "loss": 0.4948, "mean_token_accuracy": 0.8436723351478577, "num_tokens": 172049867.0, "step": 4506 }, { "epoch": 0.5733367256074291, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.239457130432129, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8591246604919434, "num_tokens": 172092176.0, "step": 4507 }, { "epoch": 0.5734639358860196, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.232674598693848, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8586311340332031, "num_tokens": 172132789.0, "step": 4508 }, { "epoch": 0.5735911461646102, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 10.233766555786133, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8726096153259277, "num_tokens": 172170536.0, "step": 4509 }, { "epoch": 0.5737183564432006, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.227790832519531, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8584462404251099, "num_tokens": 172206828.0, "step": 4510 }, { "epoch": 0.5738455667217911, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.296753883361816, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8502578735351562, "num_tokens": 172242311.0, "step": 4511 }, { "epoch": 0.5739727770003816, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.215269088745117, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.854276180267334, "num_tokens": 172275135.0, "step": 4512 }, { "epoch": 0.5740999872789722, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.229808807373047, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8670369386672974, "num_tokens": 172308682.0, "step": 4513 }, { "epoch": 0.5742271975575627, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.243264198303223, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8533134460449219, "num_tokens": 172345834.0, "step": 4514 }, { "epoch": 0.5743544078361532, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.210619926452637, "learning_rate": 1e-06, "loss": 0.5203, "mean_token_accuracy": 0.836268961429596, "num_tokens": 172387976.0, "step": 4515 }, { "epoch": 0.5744816181147436, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.259058952331543, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.858183741569519, "num_tokens": 172427447.0, "step": 4516 }, { "epoch": 0.5746088283933342, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.182199478149414, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8502579927444458, "num_tokens": 172468686.0, "step": 4517 }, { "epoch": 0.5747360386719247, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.249635696411133, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8541455268859863, "num_tokens": 172508190.0, "step": 4518 }, { "epoch": 0.5748632489505152, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.23117446899414, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8743627071380615, "num_tokens": 172552295.0, "step": 4519 }, { "epoch": 0.5749904592291057, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.238226890563965, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8633689284324646, "num_tokens": 172591065.0, "step": 4520 }, { "epoch": 0.5751176695076963, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.267964363098145, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8612337112426758, "num_tokens": 172633741.0, "step": 4521 }, { "epoch": 0.5752448797862867, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.320013046264648, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8684024810791016, "num_tokens": 172669081.0, "step": 4522 }, { "epoch": 0.5753720900648772, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.300969123840332, "learning_rate": 1e-06, "loss": 0.5153, "mean_token_accuracy": 0.836065411567688, "num_tokens": 172704956.0, "step": 4523 }, { "epoch": 0.5754993003434677, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.323115348815918, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.862293004989624, "num_tokens": 172742270.0, "step": 4524 }, { "epoch": 0.5756265106220583, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.245321273803711, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8469926118850708, "num_tokens": 172778065.0, "step": 4525 }, { "epoch": 0.5757537209006488, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.239091873168945, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8664379119873047, "num_tokens": 172815825.0, "step": 4526 }, { "epoch": 0.5758809311792393, "ewc_loss": 0.018310546875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.333534240722656, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8579757213592529, "num_tokens": 172849560.0, "step": 4527 }, { "epoch": 0.5760081414578297, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.201313972473145, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8604686260223389, "num_tokens": 172885203.0, "step": 4528 }, { "epoch": 0.5761353517364203, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.23045539855957, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8574845790863037, "num_tokens": 172922341.0, "step": 4529 }, { "epoch": 0.5762625620150108, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.283889770507812, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8643388748168945, "num_tokens": 172953363.0, "step": 4530 }, { "epoch": 0.5763897722936013, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.334311485290527, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8639975786209106, "num_tokens": 172986253.0, "step": 4531 }, { "epoch": 0.5765169825721919, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.252525329589844, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8692382574081421, "num_tokens": 173023466.0, "step": 4532 }, { "epoch": 0.5766441928507824, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.24991512298584, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.85294109582901, "num_tokens": 173063329.0, "step": 4533 }, { "epoch": 0.5767714031293728, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.271560668945312, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8645612001419067, "num_tokens": 173100676.0, "step": 4534 }, { "epoch": 0.5768986134079633, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.314889907836914, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8691317439079285, "num_tokens": 173135088.0, "step": 4535 }, { "epoch": 0.5770258236865539, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.208098411560059, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8505774736404419, "num_tokens": 173175810.0, "step": 4536 }, { "epoch": 0.5771530339651444, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.274345397949219, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8595311045646667, "num_tokens": 173220826.0, "step": 4537 }, { "epoch": 0.5772802442437349, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.23165225982666, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8693568110466003, "num_tokens": 173258411.0, "step": 4538 }, { "epoch": 0.5774074545223254, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.313629150390625, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8420623540878296, "num_tokens": 173297298.0, "step": 4539 }, { "epoch": 0.5775346648009159, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.270556449890137, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8423608541488647, "num_tokens": 173330839.0, "step": 4540 }, { "epoch": 0.5776618750795064, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.349143028259277, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8508676290512085, "num_tokens": 173370219.0, "step": 4541 }, { "epoch": 0.5777890853580969, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.21230697631836, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8587702512741089, "num_tokens": 173410425.0, "step": 4542 }, { "epoch": 0.5779162956366874, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.310866355895996, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8587823510169983, "num_tokens": 173447109.0, "step": 4543 }, { "epoch": 0.578043505915278, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.240300178527832, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8631322383880615, "num_tokens": 173490064.0, "step": 4544 }, { "epoch": 0.5781707161938685, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.312066078186035, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8607979416847229, "num_tokens": 173523791.0, "step": 4545 }, { "epoch": 0.5782979264724589, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.227782249450684, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8670530319213867, "num_tokens": 173564069.0, "step": 4546 }, { "epoch": 0.5784251367510495, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.25673770904541, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8575854301452637, "num_tokens": 173601893.0, "step": 4547 }, { "epoch": 0.57855234702964, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.366202354431152, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8680795431137085, "num_tokens": 173642299.0, "step": 4548 }, { "epoch": 0.5786795573082305, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.25084114074707, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8614797592163086, "num_tokens": 173675824.0, "step": 4549 }, { "epoch": 0.578806767586821, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.284592628479004, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8600735664367676, "num_tokens": 173712192.0, "step": 4550 }, { "epoch": 0.5789339778654116, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.28995132446289, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8663867115974426, "num_tokens": 173746949.0, "step": 4551 }, { "epoch": 0.579061188144002, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.280784606933594, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8606767654418945, "num_tokens": 173784527.0, "step": 4552 }, { "epoch": 0.5791883984225925, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.209951400756836, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8647098541259766, "num_tokens": 173820648.0, "step": 4553 }, { "epoch": 0.579315608701183, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.247977256774902, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.852449893951416, "num_tokens": 173860427.0, "step": 4554 }, { "epoch": 0.5794428189797736, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.257631301879883, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.86092209815979, "num_tokens": 173899876.0, "step": 4555 }, { "epoch": 0.5795700292583641, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 10.26278305053711, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8674890995025635, "num_tokens": 173937354.0, "step": 4556 }, { "epoch": 0.5796972395369546, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 10.234957695007324, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8486349582672119, "num_tokens": 173976039.0, "step": 4557 }, { "epoch": 0.5798244498155452, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.286882400512695, "learning_rate": 1e-06, "loss": 0.5084, "mean_token_accuracy": 0.8379517197608948, "num_tokens": 174006359.0, "step": 4558 }, { "epoch": 0.5799516600941356, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.252817153930664, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8505234122276306, "num_tokens": 174044662.0, "step": 4559 }, { "epoch": 0.5800788703727261, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.298762321472168, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8544065952301025, "num_tokens": 174077133.0, "step": 4560 }, { "epoch": 0.5802060806513166, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.21263313293457, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8541817665100098, "num_tokens": 174121873.0, "step": 4561 }, { "epoch": 0.5803332909299072, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.273730278015137, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8604295253753662, "num_tokens": 174158873.0, "step": 4562 }, { "epoch": 0.5804605012084977, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.3171968460083, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8658283948898315, "num_tokens": 174189222.0, "step": 4563 }, { "epoch": 0.5805877114870882, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.233148574829102, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8505444526672363, "num_tokens": 174229018.0, "step": 4564 }, { "epoch": 0.5807149217656786, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.258633613586426, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8591777086257935, "num_tokens": 174269215.0, "step": 4565 }, { "epoch": 0.5808421320442692, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.284584045410156, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8665592670440674, "num_tokens": 174303926.0, "step": 4566 }, { "epoch": 0.5809693423228597, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.193345069885254, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8609817028045654, "num_tokens": 174347491.0, "step": 4567 }, { "epoch": 0.5810965526014502, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.349165916442871, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8611451387405396, "num_tokens": 174384246.0, "step": 4568 }, { "epoch": 0.5812237628800407, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.277310371398926, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8582316637039185, "num_tokens": 174417081.0, "step": 4569 }, { "epoch": 0.5813509731586313, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.307472229003906, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.847967267036438, "num_tokens": 174457364.0, "step": 4570 }, { "epoch": 0.5814781834372217, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.24267292022705, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8591479063034058, "num_tokens": 174494304.0, "step": 4571 }, { "epoch": 0.5816053937158122, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.318506240844727, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8597298860549927, "num_tokens": 174528047.0, "step": 4572 }, { "epoch": 0.5817326039944027, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.202821731567383, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8635437488555908, "num_tokens": 174565990.0, "step": 4573 }, { "epoch": 0.5818598142729933, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.317388534545898, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8443026542663574, "num_tokens": 174599039.0, "step": 4574 }, { "epoch": 0.5819870245515838, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.273591995239258, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8520556092262268, "num_tokens": 174634786.0, "step": 4575 }, { "epoch": 0.5821142348301743, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.2191743850708, "learning_rate": 1e-06, "loss": 0.4749, "mean_token_accuracy": 0.8488281965255737, "num_tokens": 174677875.0, "step": 4576 }, { "epoch": 0.5822414451087647, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.298453330993652, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8523006439208984, "num_tokens": 174715421.0, "step": 4577 }, { "epoch": 0.5823686553873553, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.296354293823242, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8556239604949951, "num_tokens": 174746555.0, "step": 4578 }, { "epoch": 0.5824958656659458, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.256845474243164, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8473206758499146, "num_tokens": 174786470.0, "step": 4579 }, { "epoch": 0.5826230759445363, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.258509635925293, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8557710647583008, "num_tokens": 174827131.0, "step": 4580 }, { "epoch": 0.5827502862231269, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.263143539428711, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8521559238433838, "num_tokens": 174861391.0, "step": 4581 }, { "epoch": 0.5828774965017174, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.254727363586426, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8626290559768677, "num_tokens": 174897437.0, "step": 4582 }, { "epoch": 0.5830047067803078, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.282751083374023, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8488770723342896, "num_tokens": 174932504.0, "step": 4583 }, { "epoch": 0.5831319170588983, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.255496978759766, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.853058934211731, "num_tokens": 174974746.0, "step": 4584 }, { "epoch": 0.5832591273374889, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.259119033813477, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.852928102016449, "num_tokens": 175016155.0, "step": 4585 }, { "epoch": 0.5833863376160794, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.262642860412598, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8428859710693359, "num_tokens": 175060595.0, "step": 4586 }, { "epoch": 0.5835135478946699, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.214694023132324, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8574422597885132, "num_tokens": 175101278.0, "step": 4587 }, { "epoch": 0.5836407581732604, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.293949127197266, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8510086536407471, "num_tokens": 175139479.0, "step": 4588 }, { "epoch": 0.5837679684518509, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.244007110595703, "learning_rate": 1e-06, "loss": 0.4913, "mean_token_accuracy": 0.8408533930778503, "num_tokens": 175182376.0, "step": 4589 }, { "epoch": 0.5838951787304414, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.291402816772461, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8505312204360962, "num_tokens": 175225449.0, "step": 4590 }, { "epoch": 0.5840223890090319, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.248650550842285, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8563292026519775, "num_tokens": 175255959.0, "step": 4591 }, { "epoch": 0.5841495992876224, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.262535095214844, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8457086682319641, "num_tokens": 175292142.0, "step": 4592 }, { "epoch": 0.584276809566213, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.267012596130371, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8512594699859619, "num_tokens": 175336630.0, "step": 4593 }, { "epoch": 0.5844040198448035, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.214896202087402, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8702877759933472, "num_tokens": 175372697.0, "step": 4594 }, { "epoch": 0.5845312301233939, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.263779640197754, "learning_rate": 1e-06, "loss": 0.4895, "mean_token_accuracy": 0.8461503982543945, "num_tokens": 175412395.0, "step": 4595 }, { "epoch": 0.5846584404019844, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.259591102600098, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8563280701637268, "num_tokens": 175447295.0, "step": 4596 }, { "epoch": 0.584785650680575, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.277528762817383, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8670353889465332, "num_tokens": 175480229.0, "step": 4597 }, { "epoch": 0.5849128609591655, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.278142929077148, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8575584888458252, "num_tokens": 175523081.0, "step": 4598 }, { "epoch": 0.585040071237756, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.252767562866211, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8607741594314575, "num_tokens": 175565036.0, "step": 4599 }, { "epoch": 0.5851672815163466, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.276810646057129, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8575518727302551, "num_tokens": 175606733.0, "step": 4600 }, { "epoch": 0.585294491794937, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.268223762512207, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8728621006011963, "num_tokens": 175637663.0, "step": 4601 }, { "epoch": 0.5854217020735275, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.264373779296875, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8601930141448975, "num_tokens": 175677381.0, "step": 4602 }, { "epoch": 0.585548912352118, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.26771354675293, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.868791937828064, "num_tokens": 175713049.0, "step": 4603 }, { "epoch": 0.5856761226307086, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.25163745880127, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8484864830970764, "num_tokens": 175753681.0, "step": 4604 }, { "epoch": 0.5858033329092991, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.266852378845215, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8459320664405823, "num_tokens": 175789725.0, "step": 4605 }, { "epoch": 0.5859305431878896, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.325530052185059, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.856778621673584, "num_tokens": 175823203.0, "step": 4606 }, { "epoch": 0.5860577534664801, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.24975872039795, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8520464897155762, "num_tokens": 175860506.0, "step": 4607 }, { "epoch": 0.5861849637450706, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.276679039001465, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.873869776725769, "num_tokens": 175895080.0, "step": 4608 }, { "epoch": 0.5863121740236611, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.21235466003418, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8512710928916931, "num_tokens": 175937113.0, "step": 4609 }, { "epoch": 0.5864393843022516, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.258649826049805, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8709280490875244, "num_tokens": 175966193.0, "step": 4610 }, { "epoch": 0.5865665945808421, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.230703353881836, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8513390421867371, "num_tokens": 176003771.0, "step": 4611 }, { "epoch": 0.5866938048594327, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.243231773376465, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.864000678062439, "num_tokens": 176050069.0, "step": 4612 }, { "epoch": 0.5868210151380232, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.284290313720703, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8722118735313416, "num_tokens": 176088763.0, "step": 4613 }, { "epoch": 0.5869482254166136, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.271758079528809, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8594966530799866, "num_tokens": 176127853.0, "step": 4614 }, { "epoch": 0.5870754356952042, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.248663902282715, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8438559174537659, "num_tokens": 176173775.0, "step": 4615 }, { "epoch": 0.5872026459737947, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.263494491577148, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8613163232803345, "num_tokens": 176210878.0, "step": 4616 }, { "epoch": 0.5873298562523852, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.288182258605957, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8525451421737671, "num_tokens": 176255094.0, "step": 4617 }, { "epoch": 0.5874570665309757, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.26965045928955, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8549658060073853, "num_tokens": 176293586.0, "step": 4618 }, { "epoch": 0.5875842768095663, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.301777839660645, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8607597351074219, "num_tokens": 176329203.0, "step": 4619 }, { "epoch": 0.5877114870881567, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.284074783325195, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8746073842048645, "num_tokens": 176367984.0, "step": 4620 }, { "epoch": 0.5878386973667472, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.263433456420898, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8718060255050659, "num_tokens": 176408949.0, "step": 4621 }, { "epoch": 0.5879659076453377, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.296131134033203, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8746408224105835, "num_tokens": 176451073.0, "step": 4622 }, { "epoch": 0.5880931179239283, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.270657539367676, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8658726215362549, "num_tokens": 176490623.0, "step": 4623 }, { "epoch": 0.5882203282025188, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.291267395019531, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8458575010299683, "num_tokens": 176525771.0, "step": 4624 }, { "epoch": 0.5883475384811093, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.26718521118164, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8763688802719116, "num_tokens": 176565417.0, "step": 4625 }, { "epoch": 0.5884747487596997, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.279043197631836, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8555123805999756, "num_tokens": 176599005.0, "step": 4626 }, { "epoch": 0.5886019590382903, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.233979225158691, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8611816167831421, "num_tokens": 176637947.0, "step": 4627 }, { "epoch": 0.5887291693168808, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.27130126953125, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8483944535255432, "num_tokens": 176675241.0, "step": 4628 }, { "epoch": 0.5888563795954713, "ewc_loss": 0.0185546875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 10.257070541381836, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.851858913898468, "num_tokens": 176715375.0, "step": 4629 }, { "epoch": 0.5889835898740619, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.24862289428711, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8577735424041748, "num_tokens": 176750977.0, "step": 4630 }, { "epoch": 0.5891108001526524, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.28262710571289, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8621415495872498, "num_tokens": 176788968.0, "step": 4631 }, { "epoch": 0.5892380104312428, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.265175819396973, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8541650176048279, "num_tokens": 176824399.0, "step": 4632 }, { "epoch": 0.5893652207098333, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.259295463562012, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8540019989013672, "num_tokens": 176862053.0, "step": 4633 }, { "epoch": 0.5894924309884239, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.276423454284668, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8730736374855042, "num_tokens": 176896943.0, "step": 4634 }, { "epoch": 0.5896196412670144, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.280067443847656, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8590055704116821, "num_tokens": 176936263.0, "step": 4635 }, { "epoch": 0.5897468515456049, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.2050142288208, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8471753597259521, "num_tokens": 176976632.0, "step": 4636 }, { "epoch": 0.5898740618241954, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.317465782165527, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8639968633651733, "num_tokens": 177007036.0, "step": 4637 }, { "epoch": 0.5900012721027859, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.276062965393066, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8491683602333069, "num_tokens": 177042189.0, "step": 4638 }, { "epoch": 0.5901284823813764, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.269315719604492, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.85371994972229, "num_tokens": 177082428.0, "step": 4639 }, { "epoch": 0.5902556926599669, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.276463508605957, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8619531989097595, "num_tokens": 177124447.0, "step": 4640 }, { "epoch": 0.5903829029385574, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.253316879272461, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8743526339530945, "num_tokens": 177160516.0, "step": 4641 }, { "epoch": 0.590510113217148, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.311891555786133, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8665462136268616, "num_tokens": 177198100.0, "step": 4642 }, { "epoch": 0.5906373234957385, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.255757331848145, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8521373271942139, "num_tokens": 177243170.0, "step": 4643 }, { "epoch": 0.5907645337743289, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.240524291992188, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.856270432472229, "num_tokens": 177282976.0, "step": 4644 }, { "epoch": 0.5908917440529194, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.35129451751709, "learning_rate": 1e-06, "loss": 0.4863, "mean_token_accuracy": 0.8409497737884521, "num_tokens": 177323725.0, "step": 4645 }, { "epoch": 0.59101895433151, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.282938003540039, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8584154844284058, "num_tokens": 177352178.0, "step": 4646 }, { "epoch": 0.5911461646101005, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.310296058654785, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8561400175094604, "num_tokens": 177392322.0, "step": 4647 }, { "epoch": 0.591273374888691, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.256685256958008, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8567862510681152, "num_tokens": 177432013.0, "step": 4648 }, { "epoch": 0.5914005851672816, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.3674898147583, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.872323751449585, "num_tokens": 177473542.0, "step": 4649 }, { "epoch": 0.591527795445872, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.24417781829834, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8543521165847778, "num_tokens": 177516590.0, "step": 4650 }, { "epoch": 0.5916550057244625, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.34578800201416, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8715088963508606, "num_tokens": 177553750.0, "step": 4651 }, { "epoch": 0.591782216003053, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.324275016784668, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8516256809234619, "num_tokens": 177586093.0, "step": 4652 }, { "epoch": 0.5919094262816436, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.317851066589355, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8549706935882568, "num_tokens": 177626800.0, "step": 4653 }, { "epoch": 0.5920366365602341, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.322717666625977, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8562589287757874, "num_tokens": 177662336.0, "step": 4654 }, { "epoch": 0.5921638468388246, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.271355628967285, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8665236234664917, "num_tokens": 177706922.0, "step": 4655 }, { "epoch": 0.592291057117415, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.233841896057129, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8558413982391357, "num_tokens": 177743922.0, "step": 4656 }, { "epoch": 0.5924182673960056, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.343249320983887, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8651944398880005, "num_tokens": 177777069.0, "step": 4657 }, { "epoch": 0.5925454776745961, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.30810260772705, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8662865161895752, "num_tokens": 177810959.0, "step": 4658 }, { "epoch": 0.5926726879531866, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.269829750061035, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8494277000427246, "num_tokens": 177846366.0, "step": 4659 }, { "epoch": 0.5927998982317771, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.36347770690918, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8710340261459351, "num_tokens": 177881450.0, "step": 4660 }, { "epoch": 0.5929271085103677, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.23248291015625, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8593009114265442, "num_tokens": 177922931.0, "step": 4661 }, { "epoch": 0.5930543187889582, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.308518409729004, "learning_rate": 1e-06, "loss": 0.4867, "mean_token_accuracy": 0.8415308594703674, "num_tokens": 177962722.0, "step": 4662 }, { "epoch": 0.5931815290675486, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.316868782043457, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8564000129699707, "num_tokens": 177999956.0, "step": 4663 }, { "epoch": 0.5933087393461391, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.313796043395996, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8482235670089722, "num_tokens": 178032450.0, "step": 4664 }, { "epoch": 0.5934359496247297, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.317456245422363, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8587619066238403, "num_tokens": 178075028.0, "step": 4665 }, { "epoch": 0.5935631599033202, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.26858139038086, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8542503118515015, "num_tokens": 178108267.0, "step": 4666 }, { "epoch": 0.5936903701819107, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.283040046691895, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8591135144233704, "num_tokens": 178148091.0, "step": 4667 }, { "epoch": 0.5938175804605013, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.252427101135254, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8718329071998596, "num_tokens": 178185626.0, "step": 4668 }, { "epoch": 0.5939447907390917, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.308279991149902, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8659728765487671, "num_tokens": 178224165.0, "step": 4669 }, { "epoch": 0.5940720010176822, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.275861740112305, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8574063181877136, "num_tokens": 178264118.0, "step": 4670 }, { "epoch": 0.5941992112962727, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.325568199157715, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8701071739196777, "num_tokens": 178295634.0, "step": 4671 }, { "epoch": 0.5943264215748633, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.3341646194458, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8646066188812256, "num_tokens": 178327076.0, "step": 4672 }, { "epoch": 0.5944536318534538, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.351849555969238, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8502335548400879, "num_tokens": 178357469.0, "step": 4673 }, { "epoch": 0.5945808421320443, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.286609649658203, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8642256259918213, "num_tokens": 178393830.0, "step": 4674 }, { "epoch": 0.5947080524106347, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.323517799377441, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8584569096565247, "num_tokens": 178427560.0, "step": 4675 }, { "epoch": 0.5948352626892253, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.31320571899414, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8599522709846497, "num_tokens": 178464666.0, "step": 4676 }, { "epoch": 0.5949624729678158, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.29356861114502, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8490921258926392, "num_tokens": 178508771.0, "step": 4677 }, { "epoch": 0.5950896832464063, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.35106086730957, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8629370331764221, "num_tokens": 178548516.0, "step": 4678 }, { "epoch": 0.5952168935249968, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.287854194641113, "learning_rate": 1e-06, "loss": 0.5568, "mean_token_accuracy": 0.8236076831817627, "num_tokens": 178591276.0, "step": 4679 }, { "epoch": 0.5953441038035874, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.288122177124023, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8570401668548584, "num_tokens": 178632580.0, "step": 4680 }, { "epoch": 0.5954713140821778, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.30596923828125, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8542588949203491, "num_tokens": 178672858.0, "step": 4681 }, { "epoch": 0.5955985243607683, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.289840698242188, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8600454926490784, "num_tokens": 178706435.0, "step": 4682 }, { "epoch": 0.5957257346393589, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.288165092468262, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8556066751480103, "num_tokens": 178746309.0, "step": 4683 }, { "epoch": 0.5958529449179494, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.280281066894531, "learning_rate": 1e-06, "loss": 0.4906, "mean_token_accuracy": 0.8404816389083862, "num_tokens": 178787639.0, "step": 4684 }, { "epoch": 0.5959801551965399, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.296374320983887, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8636528253555298, "num_tokens": 178828271.0, "step": 4685 }, { "epoch": 0.5961073654751304, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.295741081237793, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8788402080535889, "num_tokens": 178862862.0, "step": 4686 }, { "epoch": 0.5962345757537209, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.265647888183594, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8702846765518188, "num_tokens": 178897969.0, "step": 4687 }, { "epoch": 0.5963617860323114, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.258641242980957, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8635143041610718, "num_tokens": 178942002.0, "step": 4688 }, { "epoch": 0.5964889963109019, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.253623008728027, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.868034839630127, "num_tokens": 178983473.0, "step": 4689 }, { "epoch": 0.5966162065894924, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.288671493530273, "learning_rate": 1e-06, "loss": 0.4744, "mean_token_accuracy": 0.8473471403121948, "num_tokens": 179022370.0, "step": 4690 }, { "epoch": 0.596743416868083, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.288505554199219, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8736850023269653, "num_tokens": 179064841.0, "step": 4691 }, { "epoch": 0.5968706271466735, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.31076717376709, "learning_rate": 1e-06, "loss": 0.5179, "mean_token_accuracy": 0.8306199312210083, "num_tokens": 179099399.0, "step": 4692 }, { "epoch": 0.5969978374252639, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.328085899353027, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8495156168937683, "num_tokens": 179133969.0, "step": 4693 }, { "epoch": 0.5971250477038544, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.355998039245605, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8559873700141907, "num_tokens": 179165456.0, "step": 4694 }, { "epoch": 0.597252257982445, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.269892692565918, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8633277416229248, "num_tokens": 179198209.0, "step": 4695 }, { "epoch": 0.5973794682610355, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.331930160522461, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.843664288520813, "num_tokens": 179233872.0, "step": 4696 }, { "epoch": 0.597506678539626, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.295382499694824, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.87940514087677, "num_tokens": 179272721.0, "step": 4697 }, { "epoch": 0.5976338888182166, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.330102920532227, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8569158315658569, "num_tokens": 179305781.0, "step": 4698 }, { "epoch": 0.597761099096807, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.315910339355469, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.863337516784668, "num_tokens": 179350049.0, "step": 4699 }, { "epoch": 0.5978883093753975, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.286430358886719, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.858318567276001, "num_tokens": 179388063.0, "step": 4700 }, { "epoch": 0.598015519653988, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.253133773803711, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8655057549476624, "num_tokens": 179427607.0, "step": 4701 }, { "epoch": 0.5981427299325786, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.340903282165527, "learning_rate": 1e-06, "loss": 0.4962, "mean_token_accuracy": 0.8427056074142456, "num_tokens": 179471456.0, "step": 4702 }, { "epoch": 0.5982699402111691, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.252995491027832, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.870517909526825, "num_tokens": 179514318.0, "step": 4703 }, { "epoch": 0.5983971504897596, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.285248756408691, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8718832731246948, "num_tokens": 179553307.0, "step": 4704 }, { "epoch": 0.59852436076835, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.274402618408203, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8505933880805969, "num_tokens": 179597288.0, "step": 4705 }, { "epoch": 0.5986515710469406, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.39515209197998, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.858269453048706, "num_tokens": 179629583.0, "step": 4706 }, { "epoch": 0.5987787813255311, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.327651023864746, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8532271981239319, "num_tokens": 179672104.0, "step": 4707 }, { "epoch": 0.5989059916041216, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.286542892456055, "learning_rate": 1e-06, "loss": 0.4883, "mean_token_accuracy": 0.8502202033996582, "num_tokens": 179712951.0, "step": 4708 }, { "epoch": 0.5990332018827121, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 10.217259407043457, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8575764298439026, "num_tokens": 179754626.0, "step": 4709 }, { "epoch": 0.5991604121613027, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.344108581542969, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.864989161491394, "num_tokens": 179790068.0, "step": 4710 }, { "epoch": 0.5992876224398932, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.340201377868652, "learning_rate": 1e-06, "loss": 0.5008, "mean_token_accuracy": 0.8398748636245728, "num_tokens": 179823936.0, "step": 4711 }, { "epoch": 0.5994148327184836, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.281349182128906, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8592848777770996, "num_tokens": 179860920.0, "step": 4712 }, { "epoch": 0.5995420429970741, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.32262134552002, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8500270843505859, "num_tokens": 179899848.0, "step": 4713 }, { "epoch": 0.5996692532756647, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.25457763671875, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8758395910263062, "num_tokens": 179934462.0, "step": 4714 }, { "epoch": 0.5997964635542552, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.285506248474121, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8596505522727966, "num_tokens": 179973955.0, "step": 4715 }, { "epoch": 0.5999236738328457, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.311551094055176, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8601170778274536, "num_tokens": 180015256.0, "step": 4716 }, { "epoch": 0.6000508841114363, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.308568954467773, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8663748502731323, "num_tokens": 180054145.0, "step": 4717 }, { "epoch": 0.6001780943900267, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.308338165283203, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8628138899803162, "num_tokens": 180093411.0, "step": 4718 }, { "epoch": 0.6003053046686172, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.34073543548584, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8685131072998047, "num_tokens": 180135799.0, "step": 4719 }, { "epoch": 0.6004325149472077, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.273950576782227, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8533985018730164, "num_tokens": 180172623.0, "step": 4720 }, { "epoch": 0.6005597252257983, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.348215103149414, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8551508188247681, "num_tokens": 180213833.0, "step": 4721 }, { "epoch": 0.6006869355043888, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.26917552947998, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.866805374622345, "num_tokens": 180254982.0, "step": 4722 }, { "epoch": 0.6008141457829793, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.369881629943848, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.867559552192688, "num_tokens": 180288544.0, "step": 4723 }, { "epoch": 0.6009413560615697, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.299910545349121, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8557214736938477, "num_tokens": 180330233.0, "step": 4724 }, { "epoch": 0.6010685663401603, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.34465503692627, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8614621162414551, "num_tokens": 180361516.0, "step": 4725 }, { "epoch": 0.6011957766187508, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.31447982788086, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8572865128517151, "num_tokens": 180404380.0, "step": 4726 }, { "epoch": 0.6013229868973413, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.348905563354492, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8520575761795044, "num_tokens": 180436916.0, "step": 4727 }, { "epoch": 0.6014501971759318, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.30314826965332, "learning_rate": 1e-06, "loss": 0.4931, "mean_token_accuracy": 0.8412606120109558, "num_tokens": 180476895.0, "step": 4728 }, { "epoch": 0.6015774074545224, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.266973495483398, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8596062660217285, "num_tokens": 180515248.0, "step": 4729 }, { "epoch": 0.6017046177331128, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.294227600097656, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.865258514881134, "num_tokens": 180554575.0, "step": 4730 }, { "epoch": 0.6018318280117033, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.32110595703125, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8498243093490601, "num_tokens": 180594485.0, "step": 4731 }, { "epoch": 0.6019590382902938, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.365368843078613, "learning_rate": 1e-06, "loss": 0.4811, "mean_token_accuracy": 0.844188928604126, "num_tokens": 180634746.0, "step": 4732 }, { "epoch": 0.6020862485688844, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.304402351379395, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8493486642837524, "num_tokens": 180674946.0, "step": 4733 }, { "epoch": 0.6022134588474749, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.357407569885254, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.859015703201294, "num_tokens": 180715538.0, "step": 4734 }, { "epoch": 0.6023406691260654, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.308732986450195, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.857641339302063, "num_tokens": 180755298.0, "step": 4735 }, { "epoch": 0.6024678794046558, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.29118537902832, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8527001142501831, "num_tokens": 180797104.0, "step": 4736 }, { "epoch": 0.6025950896832464, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.278849601745605, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8690624833106995, "num_tokens": 180839845.0, "step": 4737 }, { "epoch": 0.6027222999618369, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.312764167785645, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8737002611160278, "num_tokens": 180878849.0, "step": 4738 }, { "epoch": 0.6028495102404274, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.319755554199219, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8581206202507019, "num_tokens": 180918427.0, "step": 4739 }, { "epoch": 0.602976720519018, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.328636169433594, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8522034287452698, "num_tokens": 180956206.0, "step": 4740 }, { "epoch": 0.6031039307976085, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.328224182128906, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8647854328155518, "num_tokens": 180996945.0, "step": 4741 }, { "epoch": 0.6032311410761989, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.300942420959473, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8650268316268921, "num_tokens": 181031763.0, "step": 4742 }, { "epoch": 0.6033583513547894, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.301117897033691, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8669798374176025, "num_tokens": 181070661.0, "step": 4743 }, { "epoch": 0.60348556163338, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.313395500183105, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8453551530838013, "num_tokens": 181112769.0, "step": 4744 }, { "epoch": 0.6036127719119705, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.262065887451172, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8587957620620728, "num_tokens": 181151155.0, "step": 4745 }, { "epoch": 0.603739982190561, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.300308227539062, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8587050437927246, "num_tokens": 181188689.0, "step": 4746 }, { "epoch": 0.6038671924691515, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.264457702636719, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8635371327400208, "num_tokens": 181230031.0, "step": 4747 }, { "epoch": 0.603994402747742, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.301880836486816, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8590874671936035, "num_tokens": 181271464.0, "step": 4748 }, { "epoch": 0.6041216130263325, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.320296287536621, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8537394404411316, "num_tokens": 181312607.0, "step": 4749 }, { "epoch": 0.604248823304923, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.285999298095703, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8673774003982544, "num_tokens": 181348628.0, "step": 4750 }, { "epoch": 0.6043760335835135, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.337979316711426, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8507038950920105, "num_tokens": 181387975.0, "step": 4751 }, { "epoch": 0.6045032438621041, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.292280197143555, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8718414306640625, "num_tokens": 181424163.0, "step": 4752 }, { "epoch": 0.6046304541406946, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.308991432189941, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8489989042282104, "num_tokens": 181457002.0, "step": 4753 }, { "epoch": 0.604757664419285, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.334287643432617, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8544226884841919, "num_tokens": 181494410.0, "step": 4754 }, { "epoch": 0.6048848746978756, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.315011978149414, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.856850802898407, "num_tokens": 181534595.0, "step": 4755 }, { "epoch": 0.6050120849764661, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.297518730163574, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8653010725975037, "num_tokens": 181570654.0, "step": 4756 }, { "epoch": 0.6051392952550566, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 10.298383712768555, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8540122509002686, "num_tokens": 181608366.0, "step": 4757 }, { "epoch": 0.6052665055336471, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.326852798461914, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8468722105026245, "num_tokens": 181647168.0, "step": 4758 }, { "epoch": 0.6053937158122377, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.256729125976562, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8621358871459961, "num_tokens": 181689177.0, "step": 4759 }, { "epoch": 0.6055209260908282, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.336578369140625, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.863299548625946, "num_tokens": 181723742.0, "step": 4760 }, { "epoch": 0.6056481363694186, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.311858177185059, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8674682378768921, "num_tokens": 181761539.0, "step": 4761 }, { "epoch": 0.6057753466480091, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.327672958374023, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8567129969596863, "num_tokens": 181799674.0, "step": 4762 }, { "epoch": 0.6059025569265997, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.368383407592773, "learning_rate": 1e-06, "loss": 0.4756, "mean_token_accuracy": 0.8464193940162659, "num_tokens": 181838784.0, "step": 4763 }, { "epoch": 0.6060297672051902, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.332558631896973, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8549418449401855, "num_tokens": 181875669.0, "step": 4764 }, { "epoch": 0.6061569774837807, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.377394676208496, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.8421654105186462, "num_tokens": 181912450.0, "step": 4765 }, { "epoch": 0.6062841877623713, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.341864585876465, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8588869571685791, "num_tokens": 181947691.0, "step": 4766 }, { "epoch": 0.6064113980409617, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.355530738830566, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8628233075141907, "num_tokens": 181982364.0, "step": 4767 }, { "epoch": 0.6065386083195522, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.358441352844238, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.859100341796875, "num_tokens": 182017071.0, "step": 4768 }, { "epoch": 0.6066658185981427, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.363903045654297, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8583400249481201, "num_tokens": 182057234.0, "step": 4769 }, { "epoch": 0.6067930288767333, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.301277160644531, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8693271279335022, "num_tokens": 182092450.0, "step": 4770 }, { "epoch": 0.6069202391553238, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.428598403930664, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8511554002761841, "num_tokens": 182126398.0, "step": 4771 }, { "epoch": 0.6070474494339143, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.360222816467285, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8741368055343628, "num_tokens": 182160848.0, "step": 4772 }, { "epoch": 0.6071746597125047, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.302983283996582, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8676702380180359, "num_tokens": 182202596.0, "step": 4773 }, { "epoch": 0.6073018699910953, "ewc_loss": 0.018798828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.324316024780273, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8577658534049988, "num_tokens": 182239235.0, "step": 4774 }, { "epoch": 0.6074290802696858, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.343432426452637, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8716168999671936, "num_tokens": 182273662.0, "step": 4775 }, { "epoch": 0.6075562905482763, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.401918411254883, "learning_rate": 1e-06, "loss": 0.4941, "mean_token_accuracy": 0.8436643481254578, "num_tokens": 182311991.0, "step": 4776 }, { "epoch": 0.6076835008268668, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.35273265838623, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8607954382896423, "num_tokens": 182350722.0, "step": 4777 }, { "epoch": 0.6078107111054574, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.337176322937012, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8606967926025391, "num_tokens": 182386755.0, "step": 4778 }, { "epoch": 0.6079379213840478, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.37980842590332, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8756430149078369, "num_tokens": 182423128.0, "step": 4779 }, { "epoch": 0.6080651316626383, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.338960647583008, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.8457455635070801, "num_tokens": 182461518.0, "step": 4780 }, { "epoch": 0.6081923419412288, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.35726261138916, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8591650724411011, "num_tokens": 182495707.0, "step": 4781 }, { "epoch": 0.6083195522198194, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.326935768127441, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8492262363433838, "num_tokens": 182539058.0, "step": 4782 }, { "epoch": 0.6084467624984099, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.387852668762207, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8452244997024536, "num_tokens": 182572482.0, "step": 4783 }, { "epoch": 0.6085739727770004, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.361955642700195, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8636159896850586, "num_tokens": 182607665.0, "step": 4784 }, { "epoch": 0.6087011830555908, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.345124244689941, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.860315203666687, "num_tokens": 182650503.0, "step": 4785 }, { "epoch": 0.6088283933341814, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.32575798034668, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8698542714118958, "num_tokens": 182681662.0, "step": 4786 }, { "epoch": 0.6089556036127719, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.392184257507324, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8467862606048584, "num_tokens": 182714912.0, "step": 4787 }, { "epoch": 0.6090828138913624, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.374773979187012, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8570482730865479, "num_tokens": 182748813.0, "step": 4788 }, { "epoch": 0.609210024169953, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.400131225585938, "learning_rate": 1e-06, "loss": 0.4888, "mean_token_accuracy": 0.8428415060043335, "num_tokens": 182786123.0, "step": 4789 }, { "epoch": 0.6093372344485435, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.38810920715332, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8715047240257263, "num_tokens": 182820550.0, "step": 4790 }, { "epoch": 0.6094644447271339, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.324624061584473, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8606418371200562, "num_tokens": 182863135.0, "step": 4791 }, { "epoch": 0.6095916550057244, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.33773136138916, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8615421056747437, "num_tokens": 182905193.0, "step": 4792 }, { "epoch": 0.609718865284315, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.384598731994629, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8590577244758606, "num_tokens": 182944934.0, "step": 4793 }, { "epoch": 0.6098460755629055, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.356037139892578, "learning_rate": 1e-06, "loss": 0.4793, "mean_token_accuracy": 0.8462569117546082, "num_tokens": 182984604.0, "step": 4794 }, { "epoch": 0.609973285841496, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.416683197021484, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8578145503997803, "num_tokens": 183020421.0, "step": 4795 }, { "epoch": 0.6101004961200865, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.32326602935791, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8506633639335632, "num_tokens": 183057635.0, "step": 4796 }, { "epoch": 0.610227706398677, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.395397186279297, "learning_rate": 1e-06, "loss": 0.4777, "mean_token_accuracy": 0.8470094203948975, "num_tokens": 183094689.0, "step": 4797 }, { "epoch": 0.6103549166772675, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.418373107910156, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8646924495697021, "num_tokens": 183129118.0, "step": 4798 }, { "epoch": 0.610482126955858, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.34815502166748, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8574613928794861, "num_tokens": 183163365.0, "step": 4799 }, { "epoch": 0.6106093372344485, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.367731094360352, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8625857830047607, "num_tokens": 183199271.0, "step": 4800 }, { "epoch": 0.6107365475130391, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.354373931884766, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8784555196762085, "num_tokens": 183234523.0, "step": 4801 }, { "epoch": 0.6108637577916296, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.387304306030273, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8592991232872009, "num_tokens": 183267098.0, "step": 4802 }, { "epoch": 0.61099096807022, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.368959426879883, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8600741624832153, "num_tokens": 183306825.0, "step": 4803 }, { "epoch": 0.6111181783488105, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.348260879516602, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8587988615036011, "num_tokens": 183348712.0, "step": 4804 }, { "epoch": 0.6112453886274011, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.35417652130127, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8705777525901794, "num_tokens": 183383797.0, "step": 4805 }, { "epoch": 0.6113725989059916, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.398650169372559, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8710051774978638, "num_tokens": 183417484.0, "step": 4806 }, { "epoch": 0.6114998091845821, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.36138916015625, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8655078411102295, "num_tokens": 183453808.0, "step": 4807 }, { "epoch": 0.6116270194631727, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.327634811401367, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8665323257446289, "num_tokens": 183496693.0, "step": 4808 }, { "epoch": 0.6117542297417632, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.333638191223145, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8616787791252136, "num_tokens": 183536727.0, "step": 4809 }, { "epoch": 0.6118814400203536, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.383057594299316, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8617744445800781, "num_tokens": 183571809.0, "step": 4810 }, { "epoch": 0.6120086502989441, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.311552047729492, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8582286834716797, "num_tokens": 183610761.0, "step": 4811 }, { "epoch": 0.6121358605775347, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.350452423095703, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8573338985443115, "num_tokens": 183648304.0, "step": 4812 }, { "epoch": 0.6122630708561252, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.34406566619873, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8556227684020996, "num_tokens": 183684201.0, "step": 4813 }, { "epoch": 0.6123902811347157, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.348387718200684, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8519021272659302, "num_tokens": 183726595.0, "step": 4814 }, { "epoch": 0.6125174914133062, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.36935043334961, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.868425726890564, "num_tokens": 183762222.0, "step": 4815 }, { "epoch": 0.6126447016918967, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.38240909576416, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8712128400802612, "num_tokens": 183797167.0, "step": 4816 }, { "epoch": 0.6127719119704872, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.327674865722656, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8585078120231628, "num_tokens": 183839591.0, "step": 4817 }, { "epoch": 0.6128991222490777, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.404214859008789, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8592177033424377, "num_tokens": 183877862.0, "step": 4818 }, { "epoch": 0.6130263325276682, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.32699966430664, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.865496039390564, "num_tokens": 183915964.0, "step": 4819 }, { "epoch": 0.6131535428062588, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.38354778289795, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8564410209655762, "num_tokens": 183952823.0, "step": 4820 }, { "epoch": 0.6132807530848493, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.32196044921875, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8500598669052124, "num_tokens": 183991650.0, "step": 4821 }, { "epoch": 0.6134079633634397, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.370377540588379, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8618090152740479, "num_tokens": 184027768.0, "step": 4822 }, { "epoch": 0.6135351736420303, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.355779647827148, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.871733546257019, "num_tokens": 184063754.0, "step": 4823 }, { "epoch": 0.6136623839206208, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.394390106201172, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8531121015548706, "num_tokens": 184105286.0, "step": 4824 }, { "epoch": 0.6137895941992113, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.353992462158203, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8640588521957397, "num_tokens": 184139312.0, "step": 4825 }, { "epoch": 0.6139168044778018, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 10.360722541809082, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8590145111083984, "num_tokens": 184181541.0, "step": 4826 }, { "epoch": 0.6140440147563924, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.355846405029297, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8589516282081604, "num_tokens": 184216857.0, "step": 4827 }, { "epoch": 0.6141712250349828, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.365499496459961, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8627737760543823, "num_tokens": 184253318.0, "step": 4828 }, { "epoch": 0.6142984353135733, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.361581802368164, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8709838390350342, "num_tokens": 184298625.0, "step": 4829 }, { "epoch": 0.6144256455921638, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.422243118286133, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8583402633666992, "num_tokens": 184334569.0, "step": 4830 }, { "epoch": 0.6145528558707544, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.357830047607422, "learning_rate": 1e-06, "loss": 0.4954, "mean_token_accuracy": 0.8427003622055054, "num_tokens": 184377477.0, "step": 4831 }, { "epoch": 0.6146800661493449, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.383689880371094, "learning_rate": 1e-06, "loss": 0.4919, "mean_token_accuracy": 0.8467706441879272, "num_tokens": 184416671.0, "step": 4832 }, { "epoch": 0.6148072764279354, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.363038063049316, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8547059297561646, "num_tokens": 184453174.0, "step": 4833 }, { "epoch": 0.6149344867065258, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.421406745910645, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.866276204586029, "num_tokens": 184488313.0, "step": 4834 }, { "epoch": 0.6150616969851164, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.419092178344727, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8405750393867493, "num_tokens": 184523733.0, "step": 4835 }, { "epoch": 0.6151889072637069, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 10.397071838378906, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.85759037733078, "num_tokens": 184568399.0, "step": 4836 }, { "epoch": 0.6153161175422974, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.37718677520752, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8585767149925232, "num_tokens": 184603302.0, "step": 4837 }, { "epoch": 0.615443327820888, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.417051315307617, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8791907429695129, "num_tokens": 184637346.0, "step": 4838 }, { "epoch": 0.6155705380994785, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.399931907653809, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8663291931152344, "num_tokens": 184671687.0, "step": 4839 }, { "epoch": 0.6156977483780689, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.386459350585938, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8685359358787537, "num_tokens": 184706867.0, "step": 4840 }, { "epoch": 0.6158249586566594, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.384320259094238, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8556519746780396, "num_tokens": 184743916.0, "step": 4841 }, { "epoch": 0.61595216893525, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.391366958618164, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8681060075759888, "num_tokens": 184783247.0, "step": 4842 }, { "epoch": 0.6160793792138405, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.375604629516602, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8599693775177002, "num_tokens": 184826501.0, "step": 4843 }, { "epoch": 0.616206589492431, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.402070999145508, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8578873872756958, "num_tokens": 184868743.0, "step": 4844 }, { "epoch": 0.6163337997710215, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.356867790222168, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8557227849960327, "num_tokens": 184908992.0, "step": 4845 }, { "epoch": 0.616461010049612, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.421762466430664, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8483448624610901, "num_tokens": 184942456.0, "step": 4846 }, { "epoch": 0.6165882203282025, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.434922218322754, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8499348163604736, "num_tokens": 184975834.0, "step": 4847 }, { "epoch": 0.616715430606793, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.348384857177734, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8806021809577942, "num_tokens": 185012295.0, "step": 4848 }, { "epoch": 0.6168426408853835, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.388071060180664, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8640793561935425, "num_tokens": 185049272.0, "step": 4849 }, { "epoch": 0.6169698511639741, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.418303489685059, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8714230060577393, "num_tokens": 185081054.0, "step": 4850 }, { "epoch": 0.6170970614425646, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.353182792663574, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8561683893203735, "num_tokens": 185119809.0, "step": 4851 }, { "epoch": 0.617224271721155, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.363889694213867, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8711254000663757, "num_tokens": 185161462.0, "step": 4852 }, { "epoch": 0.6173514819997455, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.426756858825684, "learning_rate": 1e-06, "loss": 0.48, "mean_token_accuracy": 0.8512045741081238, "num_tokens": 185196238.0, "step": 4853 }, { "epoch": 0.6174786922783361, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.321770668029785, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8588587045669556, "num_tokens": 185239317.0, "step": 4854 }, { "epoch": 0.6176059025569266, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.416147232055664, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.869711697101593, "num_tokens": 185273555.0, "step": 4855 }, { "epoch": 0.6177331128355171, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.352815628051758, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8676633834838867, "num_tokens": 185309721.0, "step": 4856 }, { "epoch": 0.6178603231141077, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.379388809204102, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.852211594581604, "num_tokens": 185350846.0, "step": 4857 }, { "epoch": 0.6179875333926982, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.356451988220215, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8521658182144165, "num_tokens": 185392472.0, "step": 4858 }, { "epoch": 0.6181147436712886, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.374743461608887, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8555996417999268, "num_tokens": 185438250.0, "step": 4859 }, { "epoch": 0.6182419539498791, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.381057739257812, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8564532399177551, "num_tokens": 185479455.0, "step": 4860 }, { "epoch": 0.6183691642284697, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.392444610595703, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8564691543579102, "num_tokens": 185518719.0, "step": 4861 }, { "epoch": 0.6184963745070602, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.386834144592285, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8448437452316284, "num_tokens": 185556671.0, "step": 4862 }, { "epoch": 0.6186235847856507, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.372947692871094, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8571770191192627, "num_tokens": 185595063.0, "step": 4863 }, { "epoch": 0.6187507950642412, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.381166458129883, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8500120639801025, "num_tokens": 185635626.0, "step": 4864 }, { "epoch": 0.6188780053428317, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.419493675231934, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8478705883026123, "num_tokens": 185671687.0, "step": 4865 }, { "epoch": 0.6190052156214222, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.36485767364502, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8625656962394714, "num_tokens": 185706459.0, "step": 4866 }, { "epoch": 0.6191324259000127, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.416744232177734, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8778162598609924, "num_tokens": 185743162.0, "step": 4867 }, { "epoch": 0.6192596361786032, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.38854694366455, "learning_rate": 1e-06, "loss": 0.4892, "mean_token_accuracy": 0.8442819118499756, "num_tokens": 185784036.0, "step": 4868 }, { "epoch": 0.6193868464571938, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.426734924316406, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8555014133453369, "num_tokens": 185821407.0, "step": 4869 }, { "epoch": 0.6195140567357843, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.454732894897461, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8651686906814575, "num_tokens": 185857789.0, "step": 4870 }, { "epoch": 0.6196412670143747, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.352757453918457, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8701894283294678, "num_tokens": 185896989.0, "step": 4871 }, { "epoch": 0.6197684772929652, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.386445999145508, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.873359203338623, "num_tokens": 185938123.0, "step": 4872 }, { "epoch": 0.6198956875715558, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.394789695739746, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8600658178329468, "num_tokens": 185973440.0, "step": 4873 }, { "epoch": 0.6200228978501463, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.440865516662598, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8599214553833008, "num_tokens": 186010069.0, "step": 4874 }, { "epoch": 0.6201501081287368, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.381682395935059, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8542078137397766, "num_tokens": 186048365.0, "step": 4875 }, { "epoch": 0.6202773184073274, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.409592628479004, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8651573657989502, "num_tokens": 186085358.0, "step": 4876 }, { "epoch": 0.6204045286859178, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.400189399719238, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8637039065361023, "num_tokens": 186118476.0, "step": 4877 }, { "epoch": 0.6205317389645083, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.462191581726074, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8556816577911377, "num_tokens": 186155487.0, "step": 4878 }, { "epoch": 0.6206589492430988, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.398323059082031, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8646394610404968, "num_tokens": 186192941.0, "step": 4879 }, { "epoch": 0.6207861595216894, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.349879264831543, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8627126216888428, "num_tokens": 186233860.0, "step": 4880 }, { "epoch": 0.6209133698002799, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.414473533630371, "learning_rate": 1e-06, "loss": 0.4735, "mean_token_accuracy": 0.8485404849052429, "num_tokens": 186273081.0, "step": 4881 }, { "epoch": 0.6210405800788704, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.424773216247559, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8647864460945129, "num_tokens": 186310596.0, "step": 4882 }, { "epoch": 0.6211677903574608, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.436121940612793, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8555251359939575, "num_tokens": 186346262.0, "step": 4883 }, { "epoch": 0.6212950006360514, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.413415908813477, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8619621992111206, "num_tokens": 186386508.0, "step": 4884 }, { "epoch": 0.6214222109146419, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.380309104919434, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8555980920791626, "num_tokens": 186425184.0, "step": 4885 }, { "epoch": 0.6215494211932324, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.363190650939941, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8591241240501404, "num_tokens": 186463072.0, "step": 4886 }, { "epoch": 0.621676631471823, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.389596939086914, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8699381351470947, "num_tokens": 186504062.0, "step": 4887 }, { "epoch": 0.6218038417504135, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.35129165649414, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8501571416854858, "num_tokens": 186544564.0, "step": 4888 }, { "epoch": 0.6219310520290039, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.431978225708008, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8463665246963501, "num_tokens": 186587196.0, "step": 4889 }, { "epoch": 0.6220582623075944, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.392380714416504, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8606610298156738, "num_tokens": 186620673.0, "step": 4890 }, { "epoch": 0.622185472586185, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.354820251464844, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8623092174530029, "num_tokens": 186654028.0, "step": 4891 }, { "epoch": 0.6223126828647755, "ewc_loss": 0.01904296875, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.407246589660645, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8564175367355347, "num_tokens": 186693395.0, "step": 4892 }, { "epoch": 0.622439893143366, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.348039627075195, "learning_rate": 1e-06, "loss": 0.4936, "mean_token_accuracy": 0.8421552181243896, "num_tokens": 186734440.0, "step": 4893 }, { "epoch": 0.6225671034219565, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.410266876220703, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8620539903640747, "num_tokens": 186769637.0, "step": 4894 }, { "epoch": 0.622694313700547, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.380532264709473, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8552141189575195, "num_tokens": 186810188.0, "step": 4895 }, { "epoch": 0.6228215239791375, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.402445793151855, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8496792316436768, "num_tokens": 186847512.0, "step": 4896 }, { "epoch": 0.622948734257728, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.420083045959473, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.869705080986023, "num_tokens": 186878721.0, "step": 4897 }, { "epoch": 0.6230759445363185, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.402996063232422, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8629960417747498, "num_tokens": 186920239.0, "step": 4898 }, { "epoch": 0.6232031548149091, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.3714017868042, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8530610799789429, "num_tokens": 186964158.0, "step": 4899 }, { "epoch": 0.6233303650934996, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.41679859161377, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8741998076438904, "num_tokens": 187003176.0, "step": 4900 }, { "epoch": 0.62345757537209, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.35178279876709, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8582800030708313, "num_tokens": 187046152.0, "step": 4901 }, { "epoch": 0.6235847856506805, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.459367752075195, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8712896704673767, "num_tokens": 187086629.0, "step": 4902 }, { "epoch": 0.6237119959292711, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.43930721282959, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8538469672203064, "num_tokens": 187122856.0, "step": 4903 }, { "epoch": 0.6238392062078616, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.367300987243652, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8731964230537415, "num_tokens": 187161568.0, "step": 4904 }, { "epoch": 0.6239664164864521, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.435369491577148, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8597244024276733, "num_tokens": 187202831.0, "step": 4905 }, { "epoch": 0.6240936267650427, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.369202613830566, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8668093681335449, "num_tokens": 187239059.0, "step": 4906 }, { "epoch": 0.6242208370436332, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.406606674194336, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8534375429153442, "num_tokens": 187272940.0, "step": 4907 }, { "epoch": 0.6243480473222236, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.370317459106445, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8725046515464783, "num_tokens": 187310963.0, "step": 4908 }, { "epoch": 0.6244752576008141, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.397991180419922, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8781372308731079, "num_tokens": 187345698.0, "step": 4909 }, { "epoch": 0.6246024678794047, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.37826156616211, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8584297895431519, "num_tokens": 187384762.0, "step": 4910 }, { "epoch": 0.6247296781579952, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.394845008850098, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.849498987197876, "num_tokens": 187424288.0, "step": 4911 }, { "epoch": 0.6248568884365857, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.370552062988281, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8691989183425903, "num_tokens": 187461481.0, "step": 4912 }, { "epoch": 0.6249840987151762, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 10.381999969482422, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8739333152770996, "num_tokens": 187497690.0, "step": 4913 }, { "epoch": 0.6251113089937667, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.414913177490234, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8538649082183838, "num_tokens": 187537825.0, "step": 4914 }, { "epoch": 0.6252385192723572, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.378615379333496, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8698506355285645, "num_tokens": 187583998.0, "step": 4915 }, { "epoch": 0.6253657295509477, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.408330917358398, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8589925169944763, "num_tokens": 187621962.0, "step": 4916 }, { "epoch": 0.6254929398295382, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 10.455896377563477, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8664930462837219, "num_tokens": 187655913.0, "step": 4917 }, { "epoch": 0.6256201501081288, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.387211799621582, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8516507148742676, "num_tokens": 187692433.0, "step": 4918 }, { "epoch": 0.6257473603867193, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.444587707519531, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.872511625289917, "num_tokens": 187724311.0, "step": 4919 }, { "epoch": 0.6258745706653097, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.418412208557129, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8688763976097107, "num_tokens": 187766614.0, "step": 4920 }, { "epoch": 0.6260017809439002, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.40606689453125, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8703897595405579, "num_tokens": 187804847.0, "step": 4921 }, { "epoch": 0.6261289912224908, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.360952377319336, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8616659045219421, "num_tokens": 187842791.0, "step": 4922 }, { "epoch": 0.6262562015010813, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.490880012512207, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8510378003120422, "num_tokens": 187878304.0, "step": 4923 }, { "epoch": 0.6263834117796718, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.421931266784668, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8621004223823547, "num_tokens": 187915022.0, "step": 4924 }, { "epoch": 0.6265106220582624, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.382437705993652, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8613765239715576, "num_tokens": 187950533.0, "step": 4925 }, { "epoch": 0.6266378323368528, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.41382884979248, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8567153215408325, "num_tokens": 187989893.0, "step": 4926 }, { "epoch": 0.6267650426154433, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.3935546875, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8523463606834412, "num_tokens": 188025031.0, "step": 4927 }, { "epoch": 0.6268922528940338, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.393219947814941, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8522021174430847, "num_tokens": 188060797.0, "step": 4928 }, { "epoch": 0.6270194631726244, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.401405334472656, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.853188693523407, "num_tokens": 188096271.0, "step": 4929 }, { "epoch": 0.6271466734512149, "ewc_loss": 0.019287109375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.413285255432129, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8608525395393372, "num_tokens": 188136697.0, "step": 4930 }, { "epoch": 0.6272738837298054, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.40700626373291, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8584339618682861, "num_tokens": 188174715.0, "step": 4931 }, { "epoch": 0.6274010940083958, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.40283489227295, "learning_rate": 1e-06, "loss": 0.5163, "mean_token_accuracy": 0.838234543800354, "num_tokens": 188216379.0, "step": 4932 }, { "epoch": 0.6275283042869864, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.44511604309082, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8648544549942017, "num_tokens": 188252183.0, "step": 4933 }, { "epoch": 0.6276555145655769, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.395681381225586, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8620086908340454, "num_tokens": 188291128.0, "step": 4934 }, { "epoch": 0.6277827248441674, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.410679817199707, "learning_rate": 1e-06, "loss": 0.5022, "mean_token_accuracy": 0.8433237671852112, "num_tokens": 188334256.0, "step": 4935 }, { "epoch": 0.627909935122758, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.436514854431152, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8521118760108948, "num_tokens": 188373207.0, "step": 4936 }, { "epoch": 0.6280371454013485, "ewc_loss": 0.019775390625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.403194427490234, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8560019731521606, "num_tokens": 188414937.0, "step": 4937 }, { "epoch": 0.6281643556799389, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.387032508850098, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8559197187423706, "num_tokens": 188455398.0, "step": 4938 }, { "epoch": 0.6282915659585294, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.391096115112305, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8590607643127441, "num_tokens": 188496561.0, "step": 4939 }, { "epoch": 0.62841877623712, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.40512752532959, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8632712364196777, "num_tokens": 188531318.0, "step": 4940 }, { "epoch": 0.6285459865157105, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.412613868713379, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.850617527961731, "num_tokens": 188566807.0, "step": 4941 }, { "epoch": 0.628673196794301, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.392394065856934, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8556247353553772, "num_tokens": 188608062.0, "step": 4942 }, { "epoch": 0.6288004070728915, "ewc_loss": 0.01953125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.417484283447266, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8704444169998169, "num_tokens": 188644953.0, "step": 4943 }, { "epoch": 0.628927617351482, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.361437797546387, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8747540712356567, "num_tokens": 188687720.0, "step": 4944 }, { "epoch": 0.6290548276300725, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.41845703125, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8528159856796265, "num_tokens": 188730077.0, "step": 4945 }, { "epoch": 0.629182037908663, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.399184226989746, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8657177686691284, "num_tokens": 188768306.0, "step": 4946 }, { "epoch": 0.6293092481872535, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.41847038269043, "learning_rate": 1e-06, "loss": 0.5034, "mean_token_accuracy": 0.8387033939361572, "num_tokens": 188807480.0, "step": 4947 }, { "epoch": 0.6294364584658441, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.425264358520508, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8499559164047241, "num_tokens": 188836383.0, "step": 4948 }, { "epoch": 0.6295636687444346, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.370423316955566, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8632091879844666, "num_tokens": 188876079.0, "step": 4949 }, { "epoch": 0.629690879023025, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.38779067993164, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8570283651351929, "num_tokens": 188912936.0, "step": 4950 }, { "epoch": 0.6298180893016155, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.35744857788086, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8586594462394714, "num_tokens": 188951475.0, "step": 4951 }, { "epoch": 0.6299452995802061, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.400847434997559, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8446728587150574, "num_tokens": 188996972.0, "step": 4952 }, { "epoch": 0.6300725098587966, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.395052909851074, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8781857490539551, "num_tokens": 189030706.0, "step": 4953 }, { "epoch": 0.6301997201373871, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.358014106750488, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8816653490066528, "num_tokens": 189065638.0, "step": 4954 }, { "epoch": 0.6303269304159776, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.414692878723145, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8724116683006287, "num_tokens": 189101747.0, "step": 4955 }, { "epoch": 0.6304541406945681, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.3766450881958, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8563089370727539, "num_tokens": 189142965.0, "step": 4956 }, { "epoch": 0.6305813509731586, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.407818794250488, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8651164770126343, "num_tokens": 189177822.0, "step": 4957 }, { "epoch": 0.6307085612517491, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.41059398651123, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8543810844421387, "num_tokens": 189215710.0, "step": 4958 }, { "epoch": 0.6308357715303397, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.430135726928711, "learning_rate": 1e-06, "loss": 0.5263, "mean_token_accuracy": 0.8330731391906738, "num_tokens": 189250788.0, "step": 4959 }, { "epoch": 0.6309629818089302, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.428666114807129, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.857042670249939, "num_tokens": 189278451.0, "step": 4960 }, { "epoch": 0.6310901920875207, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.401443481445312, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8632413148880005, "num_tokens": 189316857.0, "step": 4961 }, { "epoch": 0.6312174023661112, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.430073738098145, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8706399202346802, "num_tokens": 189356600.0, "step": 4962 }, { "epoch": 0.6313446126447017, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.427738189697266, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.854297399520874, "num_tokens": 189393134.0, "step": 4963 }, { "epoch": 0.6314718229232922, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.393994331359863, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8449230194091797, "num_tokens": 189433431.0, "step": 4964 }, { "epoch": 0.6315990332018827, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.363285064697266, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8461142778396606, "num_tokens": 189473738.0, "step": 4965 }, { "epoch": 0.6317262434804732, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.413986206054688, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8736447691917419, "num_tokens": 189509930.0, "step": 4966 }, { "epoch": 0.6318534537590638, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.366125106811523, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8647412061691284, "num_tokens": 189547801.0, "step": 4967 }, { "epoch": 0.6319806640376543, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.425426483154297, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.872165322303772, "num_tokens": 189582226.0, "step": 4968 }, { "epoch": 0.6321078743162447, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.375864028930664, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.858510434627533, "num_tokens": 189622915.0, "step": 4969 }, { "epoch": 0.6322350845948352, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.403949737548828, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8678543567657471, "num_tokens": 189658182.0, "step": 4970 }, { "epoch": 0.6323622948734258, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.391434669494629, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8624448776245117, "num_tokens": 189696627.0, "step": 4971 }, { "epoch": 0.6324895051520163, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.40778636932373, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8580664396286011, "num_tokens": 189729826.0, "step": 4972 }, { "epoch": 0.6326167154306068, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.451229095458984, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8649225234985352, "num_tokens": 189771504.0, "step": 4973 }, { "epoch": 0.6327439257091974, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.350757598876953, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.877305269241333, "num_tokens": 189809220.0, "step": 4974 }, { "epoch": 0.6328711359877878, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.454704284667969, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8683797121047974, "num_tokens": 189849586.0, "step": 4975 }, { "epoch": 0.6329983462663783, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.403534889221191, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8612148761749268, "num_tokens": 189885054.0, "step": 4976 }, { "epoch": 0.6331255565449688, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.456422805786133, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8531985282897949, "num_tokens": 189918286.0, "step": 4977 }, { "epoch": 0.6332527668235594, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.39716911315918, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.8410776853561401, "num_tokens": 189956845.0, "step": 4978 }, { "epoch": 0.6333799771021499, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.472993850708008, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8768132925033569, "num_tokens": 189994827.0, "step": 4979 }, { "epoch": 0.6335071873807404, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.37005615234375, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8549931049346924, "num_tokens": 190036066.0, "step": 4980 }, { "epoch": 0.6336343976593308, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.430423736572266, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8709328770637512, "num_tokens": 190073326.0, "step": 4981 }, { "epoch": 0.6337616079379214, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.406304359436035, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8586457967758179, "num_tokens": 190109125.0, "step": 4982 }, { "epoch": 0.6338888182165119, "ewc_loss": 0.019287109375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 10.399678230285645, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8655867576599121, "num_tokens": 190148520.0, "step": 4983 }, { "epoch": 0.6340160284951024, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.391029357910156, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8629962205886841, "num_tokens": 190184272.0, "step": 4984 }, { "epoch": 0.6341432387736929, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.44465160369873, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.859052300453186, "num_tokens": 190221483.0, "step": 4985 }, { "epoch": 0.6342704490522835, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.412537574768066, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8716880083084106, "num_tokens": 190266430.0, "step": 4986 }, { "epoch": 0.6343976593308739, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.385794639587402, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8722395300865173, "num_tokens": 190308575.0, "step": 4987 }, { "epoch": 0.6345248696094644, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.411892890930176, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8706518411636353, "num_tokens": 190344584.0, "step": 4988 }, { "epoch": 0.6346520798880549, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.461137771606445, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8677866458892822, "num_tokens": 190376497.0, "step": 4989 }, { "epoch": 0.6347792901666455, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.419477462768555, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8492363095283508, "num_tokens": 190416879.0, "step": 4990 }, { "epoch": 0.634906500445236, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.376025199890137, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.880403995513916, "num_tokens": 190454684.0, "step": 4991 }, { "epoch": 0.6350337107238265, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.420210838317871, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8528485894203186, "num_tokens": 190490524.0, "step": 4992 }, { "epoch": 0.635160921002417, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.452301979064941, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.867275595664978, "num_tokens": 190526558.0, "step": 4993 }, { "epoch": 0.6352881312810075, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.416824340820312, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8610668182373047, "num_tokens": 190571206.0, "step": 4994 }, { "epoch": 0.635415341559598, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.433411598205566, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.848936915397644, "num_tokens": 190611480.0, "step": 4995 }, { "epoch": 0.6355425518381885, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.429689407348633, "learning_rate": 1e-06, "loss": 0.512, "mean_token_accuracy": 0.8331106305122375, "num_tokens": 190647699.0, "step": 4996 }, { "epoch": 0.6356697621167791, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.406261444091797, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8607403635978699, "num_tokens": 190684863.0, "step": 4997 }, { "epoch": 0.6357969723953696, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.379541397094727, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8465648293495178, "num_tokens": 190725006.0, "step": 4998 }, { "epoch": 0.63592418267396, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.460290908813477, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8576104640960693, "num_tokens": 190757431.0, "step": 4999 }, { "epoch": 0.6360513929525505, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.400683403015137, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8656235933303833, "num_tokens": 190793118.0, "step": 5000 }, { "epoch": 0.6361786032311411, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 10.381893157958984, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8705817461013794, "num_tokens": 190837105.0, "step": 5001 }, { "epoch": 0.6363058135097316, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.4283447265625, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8550164699554443, "num_tokens": 190875372.0, "step": 5002 }, { "epoch": 0.6364330237883221, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.421693801879883, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8656447529792786, "num_tokens": 190921307.0, "step": 5003 }, { "epoch": 0.6365602340669126, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.429811477661133, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8609808683395386, "num_tokens": 190967919.0, "step": 5004 }, { "epoch": 0.6366874443455031, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.437053680419922, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8592867255210876, "num_tokens": 191005668.0, "step": 5005 }, { "epoch": 0.6368146546240936, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.453227043151855, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8501237630844116, "num_tokens": 191044455.0, "step": 5006 }, { "epoch": 0.6369418649026841, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.394901275634766, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8657097816467285, "num_tokens": 191086369.0, "step": 5007 }, { "epoch": 0.6370690751812746, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.403236389160156, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8562185168266296, "num_tokens": 191127070.0, "step": 5008 }, { "epoch": 0.6371962854598652, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.447075843811035, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8752310276031494, "num_tokens": 191163623.0, "step": 5009 }, { "epoch": 0.6373234957384557, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.453581809997559, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8568601608276367, "num_tokens": 191196790.0, "step": 5010 }, { "epoch": 0.6374507060170462, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 10.483135223388672, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8583102226257324, "num_tokens": 191230265.0, "step": 5011 }, { "epoch": 0.6375779162956366, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 10.419534683227539, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8520411849021912, "num_tokens": 191266781.0, "step": 5012 }, { "epoch": 0.6377051265742272, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.418717384338379, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8508647084236145, "num_tokens": 191305869.0, "step": 5013 }, { "epoch": 0.6378323368528177, "ewc_loss": 0.01953125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.460785865783691, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8496309518814087, "num_tokens": 191343498.0, "step": 5014 }, { "epoch": 0.6379595471314082, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.421083450317383, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8458564877510071, "num_tokens": 191383260.0, "step": 5015 }, { "epoch": 0.6380867574099988, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.475919723510742, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.851798415184021, "num_tokens": 191417350.0, "step": 5016 }, { "epoch": 0.6382139676885893, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.44265079498291, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8613249659538269, "num_tokens": 191454336.0, "step": 5017 }, { "epoch": 0.6383411779671797, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.448567390441895, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8615481853485107, "num_tokens": 191489977.0, "step": 5018 }, { "epoch": 0.6384683882457702, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.453402519226074, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.8508766293525696, "num_tokens": 191524622.0, "step": 5019 }, { "epoch": 0.6385955985243608, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.421783447265625, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8494434356689453, "num_tokens": 191565389.0, "step": 5020 }, { "epoch": 0.6387228088029513, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.38278865814209, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8795386552810669, "num_tokens": 191604640.0, "step": 5021 }, { "epoch": 0.6388500190815418, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.394635200500488, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8592968583106995, "num_tokens": 191651029.0, "step": 5022 }, { "epoch": 0.6389772293601323, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.40143871307373, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8608493208885193, "num_tokens": 191690524.0, "step": 5023 }, { "epoch": 0.6391044396387228, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.450570106506348, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.856042742729187, "num_tokens": 191733992.0, "step": 5024 }, { "epoch": 0.6392316499173133, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.432112693786621, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.871431827545166, "num_tokens": 191769717.0, "step": 5025 }, { "epoch": 0.6393588601959038, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.511491775512695, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8457940816879272, "num_tokens": 191813334.0, "step": 5026 }, { "epoch": 0.6394860704744944, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.434800148010254, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8619843125343323, "num_tokens": 191848295.0, "step": 5027 }, { "epoch": 0.6396132807530849, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.438937187194824, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8533266186714172, "num_tokens": 191885329.0, "step": 5028 }, { "epoch": 0.6397404910316754, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.438007354736328, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8732398748397827, "num_tokens": 191918070.0, "step": 5029 }, { "epoch": 0.6398677013102658, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.433815956115723, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8640667200088501, "num_tokens": 191966084.0, "step": 5030 }, { "epoch": 0.6399949115888564, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.460894584655762, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8593308925628662, "num_tokens": 192004907.0, "step": 5031 }, { "epoch": 0.6401221218674469, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.416963577270508, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8424816131591797, "num_tokens": 192042942.0, "step": 5032 }, { "epoch": 0.6402493321460374, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.46630573272705, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8604534864425659, "num_tokens": 192081133.0, "step": 5033 }, { "epoch": 0.6403765424246279, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.431538581848145, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8603986501693726, "num_tokens": 192120171.0, "step": 5034 }, { "epoch": 0.6405037527032185, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.424966812133789, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8619440793991089, "num_tokens": 192159583.0, "step": 5035 }, { "epoch": 0.6406309629818089, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.421103477478027, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8580561876296997, "num_tokens": 192199513.0, "step": 5036 }, { "epoch": 0.6407581732603994, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.526189804077148, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8625901341438293, "num_tokens": 192228352.0, "step": 5037 }, { "epoch": 0.6408853835389899, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.41297721862793, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8549404144287109, "num_tokens": 192264480.0, "step": 5038 }, { "epoch": 0.6410125938175805, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.451939582824707, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8584891557693481, "num_tokens": 192303307.0, "step": 5039 }, { "epoch": 0.641139804096171, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.453455924987793, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8626450300216675, "num_tokens": 192342064.0, "step": 5040 }, { "epoch": 0.6412670143747615, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.476940155029297, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8680397272109985, "num_tokens": 192379297.0, "step": 5041 }, { "epoch": 0.6413942246533519, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.472663879394531, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8631563186645508, "num_tokens": 192416087.0, "step": 5042 }, { "epoch": 0.6415214349319425, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.433612823486328, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8650333285331726, "num_tokens": 192454421.0, "step": 5043 }, { "epoch": 0.641648645210533, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.463047981262207, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8569765686988831, "num_tokens": 192493860.0, "step": 5044 }, { "epoch": 0.6417758554891235, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.448803901672363, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8515769839286804, "num_tokens": 192533672.0, "step": 5045 }, { "epoch": 0.641903065767714, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.495850563049316, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8479717969894409, "num_tokens": 192569569.0, "step": 5046 }, { "epoch": 0.6420302760463046, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.462653160095215, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8662107586860657, "num_tokens": 192610716.0, "step": 5047 }, { "epoch": 0.642157486324895, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.504980087280273, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8541055917739868, "num_tokens": 192648040.0, "step": 5048 }, { "epoch": 0.6422846966034855, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.506521224975586, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8576871156692505, "num_tokens": 192690063.0, "step": 5049 }, { "epoch": 0.6424119068820761, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.473910331726074, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8551706671714783, "num_tokens": 192732067.0, "step": 5050 }, { "epoch": 0.6425391171606666, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.492660522460938, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8699331879615784, "num_tokens": 192768598.0, "step": 5051 }, { "epoch": 0.6426663274392571, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.517658233642578, "learning_rate": 1e-06, "loss": 0.5256, "mean_token_accuracy": 0.8320509195327759, "num_tokens": 192805420.0, "step": 5052 }, { "epoch": 0.6427935377178476, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.41653060913086, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8556592464447021, "num_tokens": 192852082.0, "step": 5053 }, { "epoch": 0.6429207479964381, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.485245704650879, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8614521622657776, "num_tokens": 192892846.0, "step": 5054 }, { "epoch": 0.6430479582750286, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.45530891418457, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8783167004585266, "num_tokens": 192932091.0, "step": 5055 }, { "epoch": 0.6431751685536191, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.544933319091797, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.862363338470459, "num_tokens": 192960628.0, "step": 5056 }, { "epoch": 0.6433023788322096, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.463765144348145, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8535922765731812, "num_tokens": 193002075.0, "step": 5057 }, { "epoch": 0.6434295891108002, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.485783576965332, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8490880727767944, "num_tokens": 193039720.0, "step": 5058 }, { "epoch": 0.6435567993893907, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.502409934997559, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8599460124969482, "num_tokens": 193072806.0, "step": 5059 }, { "epoch": 0.6436840096679812, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.432286262512207, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.868193507194519, "num_tokens": 193108478.0, "step": 5060 }, { "epoch": 0.6438112199465716, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.48093318939209, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8638765811920166, "num_tokens": 193150663.0, "step": 5061 }, { "epoch": 0.6439384302251622, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.42141342163086, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8636608123779297, "num_tokens": 193191499.0, "step": 5062 }, { "epoch": 0.6440656405037527, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.486227989196777, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8506749868392944, "num_tokens": 193236103.0, "step": 5063 }, { "epoch": 0.6441928507823432, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.508216857910156, "learning_rate": 1e-06, "loss": 0.5043, "mean_token_accuracy": 0.8380489945411682, "num_tokens": 193274095.0, "step": 5064 }, { "epoch": 0.6443200610609338, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.45787525177002, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8661850690841675, "num_tokens": 193316988.0, "step": 5065 }, { "epoch": 0.6444472713395243, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.418061256408691, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8584398031234741, "num_tokens": 193357430.0, "step": 5066 }, { "epoch": 0.6445744816181147, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.461774826049805, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.867867112159729, "num_tokens": 193390392.0, "step": 5067 }, { "epoch": 0.6447016918967052, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.420207977294922, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8572569489479065, "num_tokens": 193430961.0, "step": 5068 }, { "epoch": 0.6448289021752958, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.49778938293457, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8574435710906982, "num_tokens": 193473767.0, "step": 5069 }, { "epoch": 0.6449561124538863, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.486800193786621, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8480864763259888, "num_tokens": 193511375.0, "step": 5070 }, { "epoch": 0.6450833227324768, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.43027114868164, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8704012632369995, "num_tokens": 193551296.0, "step": 5071 }, { "epoch": 0.6452105330110673, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.437772750854492, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8637588620185852, "num_tokens": 193596484.0, "step": 5072 }, { "epoch": 0.6453377432896578, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.445549011230469, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8656591176986694, "num_tokens": 193637513.0, "step": 5073 }, { "epoch": 0.6454649535682483, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.447196006774902, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8542289733886719, "num_tokens": 193672808.0, "step": 5074 }, { "epoch": 0.6455921638468388, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.456426620483398, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.87481689453125, "num_tokens": 193713018.0, "step": 5075 }, { "epoch": 0.6457193741254293, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.448185920715332, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8620480298995972, "num_tokens": 193757649.0, "step": 5076 }, { "epoch": 0.6458465844040199, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.479713439941406, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8565009236335754, "num_tokens": 193792390.0, "step": 5077 }, { "epoch": 0.6459737946826104, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.47439956665039, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8670065402984619, "num_tokens": 193827721.0, "step": 5078 }, { "epoch": 0.6461010049612008, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.522953033447266, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8518280386924744, "num_tokens": 193867217.0, "step": 5079 }, { "epoch": 0.6462282152397913, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.498908996582031, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8675466775894165, "num_tokens": 193900128.0, "step": 5080 }, { "epoch": 0.6463554255183819, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.549345970153809, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8703312277793884, "num_tokens": 193935744.0, "step": 5081 }, { "epoch": 0.6464826357969724, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.459039688110352, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8567774891853333, "num_tokens": 193977344.0, "step": 5082 }, { "epoch": 0.6466098460755629, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.533792495727539, "learning_rate": 1e-06, "loss": 0.5177, "mean_token_accuracy": 0.8334190845489502, "num_tokens": 194014295.0, "step": 5083 }, { "epoch": 0.6467370563541535, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.465784072875977, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8453859090805054, "num_tokens": 194055245.0, "step": 5084 }, { "epoch": 0.6468642666327439, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.489093780517578, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8527935147285461, "num_tokens": 194089313.0, "step": 5085 }, { "epoch": 0.6469914769113344, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.473858833312988, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.865790843963623, "num_tokens": 194128293.0, "step": 5086 }, { "epoch": 0.6471186871899249, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.487894058227539, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8525875806808472, "num_tokens": 194166632.0, "step": 5087 }, { "epoch": 0.6472458974685155, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.457630157470703, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8460382223129272, "num_tokens": 194209176.0, "step": 5088 }, { "epoch": 0.647373107747106, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.49655818939209, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8504459261894226, "num_tokens": 194245617.0, "step": 5089 }, { "epoch": 0.6475003180256965, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.424415588378906, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8659195303916931, "num_tokens": 194285983.0, "step": 5090 }, { "epoch": 0.6476275283042869, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.511898040771484, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8558818101882935, "num_tokens": 194317088.0, "step": 5091 }, { "epoch": 0.6477547385828775, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.485920906066895, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8640052676200867, "num_tokens": 194354025.0, "step": 5092 }, { "epoch": 0.647881948861468, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 10.444144248962402, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.863553524017334, "num_tokens": 194388539.0, "step": 5093 }, { "epoch": 0.6480091591400585, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.468608856201172, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8618726134300232, "num_tokens": 194424518.0, "step": 5094 }, { "epoch": 0.648136369418649, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.470210075378418, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8556029796600342, "num_tokens": 194463310.0, "step": 5095 }, { "epoch": 0.6482635796972396, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 10.441924095153809, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.859139621257782, "num_tokens": 194499911.0, "step": 5096 }, { "epoch": 0.64839078997583, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.491427421569824, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8498876094818115, "num_tokens": 194536731.0, "step": 5097 }, { "epoch": 0.6485180002544205, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.440450668334961, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8495092391967773, "num_tokens": 194574014.0, "step": 5098 }, { "epoch": 0.648645210533011, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.467448234558105, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8511852622032166, "num_tokens": 194611944.0, "step": 5099 }, { "epoch": 0.6487724208116016, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.438804626464844, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8689510226249695, "num_tokens": 194649734.0, "step": 5100 }, { "epoch": 0.6488996310901921, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.460949897766113, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8676480054855347, "num_tokens": 194688344.0, "step": 5101 }, { "epoch": 0.6490268413687826, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.424874305725098, "learning_rate": 1e-06, "loss": 0.4931, "mean_token_accuracy": 0.8466785550117493, "num_tokens": 194728582.0, "step": 5102 }, { "epoch": 0.649154051647373, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.4459228515625, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8599401712417603, "num_tokens": 194762753.0, "step": 5103 }, { "epoch": 0.6492812619259636, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.401822090148926, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8728598356246948, "num_tokens": 194802325.0, "step": 5104 }, { "epoch": 0.6494084722045541, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.474132537841797, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8579995036125183, "num_tokens": 194846349.0, "step": 5105 }, { "epoch": 0.6495356824831446, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.424163818359375, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8609020709991455, "num_tokens": 194880334.0, "step": 5106 }, { "epoch": 0.6496628927617352, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.495413780212402, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8508902788162231, "num_tokens": 194919343.0, "step": 5107 }, { "epoch": 0.6497901030403257, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.473485946655273, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8600121736526489, "num_tokens": 194958800.0, "step": 5108 }, { "epoch": 0.6499173133189162, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.454724311828613, "learning_rate": 1e-06, "loss": 0.5004, "mean_token_accuracy": 0.8334515690803528, "num_tokens": 194997249.0, "step": 5109 }, { "epoch": 0.6500445235975066, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.519796371459961, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8620840311050415, "num_tokens": 195035879.0, "step": 5110 }, { "epoch": 0.6501717338760972, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.486457824707031, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8543005585670471, "num_tokens": 195075925.0, "step": 5111 }, { "epoch": 0.6502989441546877, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.530237197875977, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8620626926422119, "num_tokens": 195110185.0, "step": 5112 }, { "epoch": 0.6504261544332782, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.50091552734375, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8515195250511169, "num_tokens": 195154666.0, "step": 5113 }, { "epoch": 0.6505533647118688, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.527044296264648, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8716799020767212, "num_tokens": 195188998.0, "step": 5114 }, { "epoch": 0.6506805749904593, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.492823600769043, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8668515682220459, "num_tokens": 195224124.0, "step": 5115 }, { "epoch": 0.6508077852690497, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.53291130065918, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8604111671447754, "num_tokens": 195253601.0, "step": 5116 }, { "epoch": 0.6509349955476402, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.468461036682129, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8513320088386536, "num_tokens": 195289045.0, "step": 5117 }, { "epoch": 0.6510622058262308, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.447726249694824, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8611077070236206, "num_tokens": 195326983.0, "step": 5118 }, { "epoch": 0.6511894161048213, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.500702857971191, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8716049194335938, "num_tokens": 195360425.0, "step": 5119 }, { "epoch": 0.6513166263834118, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.472795486450195, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8487750887870789, "num_tokens": 195402648.0, "step": 5120 }, { "epoch": 0.6514438366620023, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.4576997756958, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8772178888320923, "num_tokens": 195440945.0, "step": 5121 }, { "epoch": 0.6515710469405928, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.470730781555176, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8618811368942261, "num_tokens": 195481872.0, "step": 5122 }, { "epoch": 0.6516982572191833, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.490717887878418, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8645191192626953, "num_tokens": 195525460.0, "step": 5123 }, { "epoch": 0.6518254674977738, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.529271125793457, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8621612191200256, "num_tokens": 195563522.0, "step": 5124 }, { "epoch": 0.6519526777763643, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.429826736450195, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8687735199928284, "num_tokens": 195598114.0, "step": 5125 }, { "epoch": 0.6520798880549549, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.482900619506836, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8550519943237305, "num_tokens": 195639826.0, "step": 5126 }, { "epoch": 0.6522070983335454, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.481459617614746, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8663592338562012, "num_tokens": 195682047.0, "step": 5127 }, { "epoch": 0.6523343086121358, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.503085136413574, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8644185662269592, "num_tokens": 195722453.0, "step": 5128 }, { "epoch": 0.6524615188907263, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.453177452087402, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8553609251976013, "num_tokens": 195759182.0, "step": 5129 }, { "epoch": 0.6525887291693169, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.474092483520508, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8639189004898071, "num_tokens": 195795641.0, "step": 5130 }, { "epoch": 0.6527159394479074, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.500870704650879, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8695234656333923, "num_tokens": 195831200.0, "step": 5131 }, { "epoch": 0.6528431497264979, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.431794166564941, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8628770709037781, "num_tokens": 195875110.0, "step": 5132 }, { "epoch": 0.6529703600050885, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.48875617980957, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8632643222808838, "num_tokens": 195916724.0, "step": 5133 }, { "epoch": 0.6530975702836789, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.407145500183105, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8541444540023804, "num_tokens": 195956219.0, "step": 5134 }, { "epoch": 0.6532247805622694, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.546046257019043, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8639170527458191, "num_tokens": 195991563.0, "step": 5135 }, { "epoch": 0.6533519908408599, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.428021430969238, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8671577572822571, "num_tokens": 196031864.0, "step": 5136 }, { "epoch": 0.6534792011194505, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.522456169128418, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8526414036750793, "num_tokens": 196067208.0, "step": 5137 }, { "epoch": 0.653606411398041, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.468693733215332, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8640546798706055, "num_tokens": 196103934.0, "step": 5138 }, { "epoch": 0.6537336216766315, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.530904769897461, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8548597097396851, "num_tokens": 196140886.0, "step": 5139 }, { "epoch": 0.6538608319552219, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.490339279174805, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8695474863052368, "num_tokens": 196178940.0, "step": 5140 }, { "epoch": 0.6539880422338125, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.439980506896973, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8622274994850159, "num_tokens": 196229461.0, "step": 5141 }, { "epoch": 0.654115252512403, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.506373405456543, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8732650876045227, "num_tokens": 196265793.0, "step": 5142 }, { "epoch": 0.6542424627909935, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.456367492675781, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8677048683166504, "num_tokens": 196304919.0, "step": 5143 }, { "epoch": 0.654369673069584, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.46687126159668, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.876057505607605, "num_tokens": 196344761.0, "step": 5144 }, { "epoch": 0.6544968833481746, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.425065040588379, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8747348189353943, "num_tokens": 196382625.0, "step": 5145 }, { "epoch": 0.654624093626765, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.476143836975098, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8718472123146057, "num_tokens": 196422279.0, "step": 5146 }, { "epoch": 0.6547513039053555, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.448195457458496, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8566222190856934, "num_tokens": 196463684.0, "step": 5147 }, { "epoch": 0.654878514183946, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.477660179138184, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8637457489967346, "num_tokens": 196506319.0, "step": 5148 }, { "epoch": 0.6550057244625366, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.466261863708496, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8540388941764832, "num_tokens": 196546367.0, "step": 5149 }, { "epoch": 0.6551329347411271, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.535696029663086, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8461390137672424, "num_tokens": 196576598.0, "step": 5150 }, { "epoch": 0.6552601450197176, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.501803398132324, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.854453444480896, "num_tokens": 196610633.0, "step": 5151 }, { "epoch": 0.655387355298308, "ewc_loss": 0.019775390625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.51659107208252, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.852198600769043, "num_tokens": 196646183.0, "step": 5152 }, { "epoch": 0.6555145655768986, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.447318077087402, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.869867205619812, "num_tokens": 196689454.0, "step": 5153 }, { "epoch": 0.6556417758554891, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 10.488880157470703, "learning_rate": 1e-06, "loss": 0.4764, "mean_token_accuracy": 0.8452733755111694, "num_tokens": 196731820.0, "step": 5154 }, { "epoch": 0.6557689861340796, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.498339653015137, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8512447476387024, "num_tokens": 196768542.0, "step": 5155 }, { "epoch": 0.6558961964126702, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.488506317138672, "learning_rate": 1e-06, "loss": 0.4815, "mean_token_accuracy": 0.8461170196533203, "num_tokens": 196810300.0, "step": 5156 }, { "epoch": 0.6560234066912607, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.52813720703125, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8440836668014526, "num_tokens": 196847271.0, "step": 5157 }, { "epoch": 0.6561506169698512, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.448403358459473, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8753033876419067, "num_tokens": 196880439.0, "step": 5158 }, { "epoch": 0.6562778272484416, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.534127235412598, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8539489507675171, "num_tokens": 196914911.0, "step": 5159 }, { "epoch": 0.6564050375270322, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.484957695007324, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8607571125030518, "num_tokens": 196949419.0, "step": 5160 }, { "epoch": 0.6565322478056227, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.523303031921387, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8531769514083862, "num_tokens": 196989401.0, "step": 5161 }, { "epoch": 0.6566594580842132, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.47127628326416, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8536968231201172, "num_tokens": 197028377.0, "step": 5162 }, { "epoch": 0.6567866683628037, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.513629913330078, "learning_rate": 1e-06, "loss": 0.494, "mean_token_accuracy": 0.8397489190101624, "num_tokens": 197068334.0, "step": 5163 }, { "epoch": 0.6569138786413943, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.445280075073242, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8631572723388672, "num_tokens": 197105978.0, "step": 5164 }, { "epoch": 0.6570410889199847, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.530972480773926, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8436416387557983, "num_tokens": 197148641.0, "step": 5165 }, { "epoch": 0.6571682991985752, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.49205207824707, "learning_rate": 1e-06, "loss": 0.4967, "mean_token_accuracy": 0.8414092063903809, "num_tokens": 197188820.0, "step": 5166 }, { "epoch": 0.6572955094771658, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.48985481262207, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8475659489631653, "num_tokens": 197227109.0, "step": 5167 }, { "epoch": 0.6574227197557563, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.521135330200195, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8661656975746155, "num_tokens": 197264083.0, "step": 5168 }, { "epoch": 0.6575499300343468, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.495762825012207, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8563470244407654, "num_tokens": 197301349.0, "step": 5169 }, { "epoch": 0.6576771403129373, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.472661972045898, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.8491063117980957, "num_tokens": 197348176.0, "step": 5170 }, { "epoch": 0.6578043505915278, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.46397876739502, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8526804447174072, "num_tokens": 197380490.0, "step": 5171 }, { "epoch": 0.6579315608701183, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.499490737915039, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8471153378486633, "num_tokens": 197423075.0, "step": 5172 }, { "epoch": 0.6580587711487088, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.420490264892578, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8651986122131348, "num_tokens": 197461730.0, "step": 5173 }, { "epoch": 0.6581859814272993, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.469602584838867, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8485986590385437, "num_tokens": 197505801.0, "step": 5174 }, { "epoch": 0.6583131917058899, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.46453857421875, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8776687383651733, "num_tokens": 197534643.0, "step": 5175 }, { "epoch": 0.6584404019844804, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.451178550720215, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8656938076019287, "num_tokens": 197576490.0, "step": 5176 }, { "epoch": 0.6585676122630708, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.482060432434082, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8730936050415039, "num_tokens": 197612745.0, "step": 5177 }, { "epoch": 0.6586948225416613, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.499516487121582, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8551982641220093, "num_tokens": 197653756.0, "step": 5178 }, { "epoch": 0.6588220328202519, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.447225570678711, "learning_rate": 1e-06, "loss": 0.353, "mean_token_accuracy": 0.8843051195144653, "num_tokens": 197694417.0, "step": 5179 }, { "epoch": 0.6589492430988424, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.491867065429688, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.855767548084259, "num_tokens": 197732738.0, "step": 5180 }, { "epoch": 0.6590764533774329, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.523246765136719, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8512283563613892, "num_tokens": 197759408.0, "step": 5181 }, { "epoch": 0.6592036636560235, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.517687797546387, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8626460433006287, "num_tokens": 197794334.0, "step": 5182 }, { "epoch": 0.6593308739346139, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.527314186096191, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8693791627883911, "num_tokens": 197832497.0, "step": 5183 }, { "epoch": 0.6594580842132044, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.481035232543945, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8518311977386475, "num_tokens": 197869021.0, "step": 5184 }, { "epoch": 0.6595852944917949, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.526111602783203, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8467060923576355, "num_tokens": 197906365.0, "step": 5185 }, { "epoch": 0.6597125047703855, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.460217475891113, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8626790046691895, "num_tokens": 197950725.0, "step": 5186 }, { "epoch": 0.659839715048976, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.515997886657715, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8681910037994385, "num_tokens": 197991178.0, "step": 5187 }, { "epoch": 0.6599669253275665, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.584001541137695, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.851464569568634, "num_tokens": 198030176.0, "step": 5188 }, { "epoch": 0.6600941356061569, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.505013465881348, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8547931909561157, "num_tokens": 198067169.0, "step": 5189 }, { "epoch": 0.6602213458847475, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.568658828735352, "learning_rate": 1e-06, "loss": 0.4705, "mean_token_accuracy": 0.849835216999054, "num_tokens": 198104994.0, "step": 5190 }, { "epoch": 0.660348556163338, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.494906425476074, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8647967576980591, "num_tokens": 198142791.0, "step": 5191 }, { "epoch": 0.6604757664419285, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.500831604003906, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8563035726547241, "num_tokens": 198179300.0, "step": 5192 }, { "epoch": 0.660602976720519, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.496072769165039, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8741726875305176, "num_tokens": 198220230.0, "step": 5193 }, { "epoch": 0.6607301869991096, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.576025009155273, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8487738966941833, "num_tokens": 198251067.0, "step": 5194 }, { "epoch": 0.6608573972777, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.528702735900879, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.869642972946167, "num_tokens": 198282130.0, "step": 5195 }, { "epoch": 0.6609846075562905, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.440550804138184, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8700724244117737, "num_tokens": 198325609.0, "step": 5196 }, { "epoch": 0.661111817834881, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.526152610778809, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8635565042495728, "num_tokens": 198365116.0, "step": 5197 }, { "epoch": 0.6612390281134716, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.514984130859375, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.860100507736206, "num_tokens": 198403467.0, "step": 5198 }, { "epoch": 0.6613662383920621, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.470958709716797, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.866227388381958, "num_tokens": 198440676.0, "step": 5199 }, { "epoch": 0.6614934486706526, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.51633071899414, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8619369268417358, "num_tokens": 198479082.0, "step": 5200 }, { "epoch": 0.661620658949243, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.518874168395996, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8561491966247559, "num_tokens": 198521741.0, "step": 5201 }, { "epoch": 0.6617478692278336, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.486725807189941, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8534741401672363, "num_tokens": 198562177.0, "step": 5202 }, { "epoch": 0.6618750795064241, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.503544807434082, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8608842492103577, "num_tokens": 198600726.0, "step": 5203 }, { "epoch": 0.6620022897850146, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.520265579223633, "learning_rate": 1e-06, "loss": 0.523, "mean_token_accuracy": 0.8318115472793579, "num_tokens": 198638156.0, "step": 5204 }, { "epoch": 0.6621295000636052, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.529298782348633, "learning_rate": 1e-06, "loss": 0.5077, "mean_token_accuracy": 0.8422746658325195, "num_tokens": 198680312.0, "step": 5205 }, { "epoch": 0.6622567103421957, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.477001190185547, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8777555227279663, "num_tokens": 198723886.0, "step": 5206 }, { "epoch": 0.6623839206207861, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.541976928710938, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8602135181427002, "num_tokens": 198765331.0, "step": 5207 }, { "epoch": 0.6625111308993766, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.499444961547852, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8638032078742981, "num_tokens": 198802421.0, "step": 5208 }, { "epoch": 0.6626383411779672, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.5645751953125, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8705098628997803, "num_tokens": 198843335.0, "step": 5209 }, { "epoch": 0.6627655514565577, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.521919250488281, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8668373823165894, "num_tokens": 198880925.0, "step": 5210 }, { "epoch": 0.6628927617351482, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.493072509765625, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8591099977493286, "num_tokens": 198915448.0, "step": 5211 }, { "epoch": 0.6630199720137387, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.51149845123291, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8708646297454834, "num_tokens": 198952254.0, "step": 5212 }, { "epoch": 0.6631471822923293, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.530144691467285, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8652172684669495, "num_tokens": 198990841.0, "step": 5213 }, { "epoch": 0.6632743925709197, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.50421142578125, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8625648021697998, "num_tokens": 199029383.0, "step": 5214 }, { "epoch": 0.6634016028495102, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.613874435424805, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8598126173019409, "num_tokens": 199066731.0, "step": 5215 }, { "epoch": 0.6635288131281007, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.525056838989258, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8728432059288025, "num_tokens": 199108752.0, "step": 5216 }, { "epoch": 0.6636560234066913, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.529936790466309, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8640210628509521, "num_tokens": 199146438.0, "step": 5217 }, { "epoch": 0.6637832336852818, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.495942115783691, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8624349236488342, "num_tokens": 199186367.0, "step": 5218 }, { "epoch": 0.6639104439638723, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.548148155212402, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8741824626922607, "num_tokens": 199225785.0, "step": 5219 }, { "epoch": 0.6640376542424627, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.468304634094238, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8639155626296997, "num_tokens": 199262574.0, "step": 5220 }, { "epoch": 0.6641648645210533, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.533808708190918, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8658563494682312, "num_tokens": 199299093.0, "step": 5221 }, { "epoch": 0.6642920747996438, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.494905471801758, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8663803339004517, "num_tokens": 199337446.0, "step": 5222 }, { "epoch": 0.6644192850782343, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.52431869506836, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8625971674919128, "num_tokens": 199377321.0, "step": 5223 }, { "epoch": 0.6645464953568249, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.49243450164795, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8713600635528564, "num_tokens": 199413057.0, "step": 5224 }, { "epoch": 0.6646737056354154, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.528218269348145, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8680292367935181, "num_tokens": 199452730.0, "step": 5225 }, { "epoch": 0.6648009159140058, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.498371124267578, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8720091581344604, "num_tokens": 199487860.0, "step": 5226 }, { "epoch": 0.6649281261925963, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.513771057128906, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8678067922592163, "num_tokens": 199532790.0, "step": 5227 }, { "epoch": 0.6650553364711869, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 10.524699211120605, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8803838491439819, "num_tokens": 199561072.0, "step": 5228 }, { "epoch": 0.6651825467497774, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.463428497314453, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8665703535079956, "num_tokens": 199595820.0, "step": 5229 }, { "epoch": 0.6653097570283679, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.497251510620117, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8490657806396484, "num_tokens": 199640101.0, "step": 5230 }, { "epoch": 0.6654369673069584, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.465703010559082, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8573682904243469, "num_tokens": 199680287.0, "step": 5231 }, { "epoch": 0.6655641775855489, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.497698783874512, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8602755069732666, "num_tokens": 199721958.0, "step": 5232 }, { "epoch": 0.6656913878641394, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.504254341125488, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8675668835639954, "num_tokens": 199756021.0, "step": 5233 }, { "epoch": 0.6658185981427299, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.519919395446777, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8662430047988892, "num_tokens": 199793586.0, "step": 5234 }, { "epoch": 0.6659458084213205, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.515963554382324, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8479975461959839, "num_tokens": 199835728.0, "step": 5235 }, { "epoch": 0.666073018699911, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.470108032226562, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8706697225570679, "num_tokens": 199873348.0, "step": 5236 }, { "epoch": 0.6662002289785015, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.55497932434082, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8610565662384033, "num_tokens": 199908223.0, "step": 5237 }, { "epoch": 0.6663274392570919, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.526820182800293, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8581863641738892, "num_tokens": 199947017.0, "step": 5238 }, { "epoch": 0.6664546495356825, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.507328987121582, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.869811475276947, "num_tokens": 199984806.0, "step": 5239 }, { "epoch": 0.666581859814273, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.534177780151367, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.855938196182251, "num_tokens": 200022600.0, "step": 5240 }, { "epoch": 0.6667090700928635, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.482767105102539, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8722042441368103, "num_tokens": 200058407.0, "step": 5241 }, { "epoch": 0.666836280371454, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.543153762817383, "learning_rate": 1e-06, "loss": 0.5128, "mean_token_accuracy": 0.8340420722961426, "num_tokens": 200101871.0, "step": 5242 }, { "epoch": 0.6669634906500446, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.536231994628906, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8614456653594971, "num_tokens": 200138501.0, "step": 5243 }, { "epoch": 0.667090700928635, "ewc_loss": 0.02001953125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 10.469511985778809, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8597840070724487, "num_tokens": 200176639.0, "step": 5244 }, { "epoch": 0.6672179112072255, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.522786140441895, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8566207885742188, "num_tokens": 200211873.0, "step": 5245 }, { "epoch": 0.667345121485816, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.511456489562988, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8572745323181152, "num_tokens": 200246596.0, "step": 5246 }, { "epoch": 0.6674723317644066, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.533184051513672, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8765520453453064, "num_tokens": 200282761.0, "step": 5247 }, { "epoch": 0.6675995420429971, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.500751495361328, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.8470945358276367, "num_tokens": 200321821.0, "step": 5248 }, { "epoch": 0.6677267523215876, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.51195240020752, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.844659149646759, "num_tokens": 200369066.0, "step": 5249 }, { "epoch": 0.667853962600178, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.480094909667969, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8532928824424744, "num_tokens": 200408067.0, "step": 5250 }, { "epoch": 0.6679811728787686, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.505230903625488, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8776394128799438, "num_tokens": 200446693.0, "step": 5251 }, { "epoch": 0.6681083831573591, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.49953842163086, "learning_rate": 1e-06, "loss": 0.4974, "mean_token_accuracy": 0.8428407907485962, "num_tokens": 200485340.0, "step": 5252 }, { "epoch": 0.6682355934359496, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.523917198181152, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8581965565681458, "num_tokens": 200521978.0, "step": 5253 }, { "epoch": 0.6683628037145402, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.532252311706543, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8627532124519348, "num_tokens": 200560604.0, "step": 5254 }, { "epoch": 0.6684900139931307, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.5110445022583, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.85755854845047, "num_tokens": 200601718.0, "step": 5255 }, { "epoch": 0.6686172242717211, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.501895904541016, "learning_rate": 1e-06, "loss": 0.5054, "mean_token_accuracy": 0.8361985087394714, "num_tokens": 200646186.0, "step": 5256 }, { "epoch": 0.6687444345503116, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.496077537536621, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.863540530204773, "num_tokens": 200684849.0, "step": 5257 }, { "epoch": 0.6688716448289022, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.5272855758667, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8668643236160278, "num_tokens": 200719542.0, "step": 5258 }, { "epoch": 0.6689988551074927, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.47265911102295, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8623191714286804, "num_tokens": 200756212.0, "step": 5259 }, { "epoch": 0.6691260653860832, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.528743743896484, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8631387948989868, "num_tokens": 200791184.0, "step": 5260 }, { "epoch": 0.6692532756646737, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.515093803405762, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8450644016265869, "num_tokens": 200823966.0, "step": 5261 }, { "epoch": 0.6693804859432643, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.521039009094238, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8659590482711792, "num_tokens": 200862718.0, "step": 5262 }, { "epoch": 0.6695076962218547, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.480300903320312, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8555854558944702, "num_tokens": 200901755.0, "step": 5263 }, { "epoch": 0.6696349065004452, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.542869567871094, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8719910979270935, "num_tokens": 200935788.0, "step": 5264 }, { "epoch": 0.6697621167790357, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.493343353271484, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8755244016647339, "num_tokens": 200975159.0, "step": 5265 }, { "epoch": 0.6698893270576263, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.543354988098145, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8731955289840698, "num_tokens": 201010813.0, "step": 5266 }, { "epoch": 0.6700165373362168, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.560172080993652, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8590793013572693, "num_tokens": 201051937.0, "step": 5267 }, { "epoch": 0.6701437476148073, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.511651039123535, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.866734504699707, "num_tokens": 201090926.0, "step": 5268 }, { "epoch": 0.6702709578933977, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.510148048400879, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8418114185333252, "num_tokens": 201126086.0, "step": 5269 }, { "epoch": 0.6703981681719883, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.548197746276855, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.850662887096405, "num_tokens": 201165933.0, "step": 5270 }, { "epoch": 0.6705253784505788, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.586492538452148, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8640869855880737, "num_tokens": 201198794.0, "step": 5271 }, { "epoch": 0.6706525887291693, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.516681671142578, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8526319861412048, "num_tokens": 201234081.0, "step": 5272 }, { "epoch": 0.6707797990077599, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.571016311645508, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8571504950523376, "num_tokens": 201271796.0, "step": 5273 }, { "epoch": 0.6709070092863504, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.513907432556152, "learning_rate": 1e-06, "loss": 0.4825, "mean_token_accuracy": 0.8428292274475098, "num_tokens": 201309526.0, "step": 5274 }, { "epoch": 0.6710342195649408, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.545741081237793, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.867988109588623, "num_tokens": 201344015.0, "step": 5275 }, { "epoch": 0.6711614298435313, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.529193878173828, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.861964225769043, "num_tokens": 201385201.0, "step": 5276 }, { "epoch": 0.6712886401221219, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.533781051635742, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8664497137069702, "num_tokens": 201422370.0, "step": 5277 }, { "epoch": 0.6714158504007124, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.478263854980469, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8462162613868713, "num_tokens": 201467085.0, "step": 5278 }, { "epoch": 0.6715430606793029, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.576044082641602, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8752747774124146, "num_tokens": 201503432.0, "step": 5279 }, { "epoch": 0.6716702709578934, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.504405975341797, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8680249452590942, "num_tokens": 201541358.0, "step": 5280 }, { "epoch": 0.6717974812364839, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.514527320861816, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8581470251083374, "num_tokens": 201580235.0, "step": 5281 }, { "epoch": 0.6719246915150744, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.52795124053955, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8613408803939819, "num_tokens": 201627703.0, "step": 5282 }, { "epoch": 0.6720519017936649, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.544933319091797, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8641409873962402, "num_tokens": 201667182.0, "step": 5283 }, { "epoch": 0.6721791120722554, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.555190086364746, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8618335723876953, "num_tokens": 201699876.0, "step": 5284 }, { "epoch": 0.672306322350846, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.55990982055664, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8520183563232422, "num_tokens": 201732770.0, "step": 5285 }, { "epoch": 0.6724335326294365, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.506951332092285, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8630073666572571, "num_tokens": 201771388.0, "step": 5286 }, { "epoch": 0.6725607429080269, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.511584281921387, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8560761213302612, "num_tokens": 201809283.0, "step": 5287 }, { "epoch": 0.6726879531866174, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.544133186340332, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8679084777832031, "num_tokens": 201854950.0, "step": 5288 }, { "epoch": 0.672815163465208, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.555994033813477, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8650677800178528, "num_tokens": 201895638.0, "step": 5289 }, { "epoch": 0.6729423737437985, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.585057258605957, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8481549024581909, "num_tokens": 201928243.0, "step": 5290 }, { "epoch": 0.673069584022389, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.565120697021484, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8608406782150269, "num_tokens": 201964687.0, "step": 5291 }, { "epoch": 0.6731967943009796, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.541126251220703, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8742148876190186, "num_tokens": 201997166.0, "step": 5292 }, { "epoch": 0.67332400457957, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.584589958190918, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8633111715316772, "num_tokens": 202030771.0, "step": 5293 }, { "epoch": 0.6734512148581605, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.563765525817871, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8533127307891846, "num_tokens": 202077280.0, "step": 5294 }, { "epoch": 0.673578425136751, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.549712181091309, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8531164526939392, "num_tokens": 202115273.0, "step": 5295 }, { "epoch": 0.6737056354153416, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.582657814025879, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8713175654411316, "num_tokens": 202145347.0, "step": 5296 }, { "epoch": 0.6738328456939321, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.573081970214844, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8660436272621155, "num_tokens": 202184621.0, "step": 5297 }, { "epoch": 0.6739600559725226, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.477473258972168, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8639943599700928, "num_tokens": 202230099.0, "step": 5298 }, { "epoch": 0.674087266251113, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.610523223876953, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8552619218826294, "num_tokens": 202267872.0, "step": 5299 }, { "epoch": 0.6742144765297036, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.537956237792969, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8591577410697937, "num_tokens": 202305517.0, "step": 5300 }, { "epoch": 0.6743416868082941, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.565549850463867, "learning_rate": 1e-06, "loss": 0.4736, "mean_token_accuracy": 0.847784698009491, "num_tokens": 202340671.0, "step": 5301 }, { "epoch": 0.6744688970868846, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.57015323638916, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8697126507759094, "num_tokens": 202379552.0, "step": 5302 }, { "epoch": 0.6745961073654752, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.51192855834961, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8567167520523071, "num_tokens": 202421404.0, "step": 5303 }, { "epoch": 0.6747233176440657, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.537801742553711, "learning_rate": 1e-06, "loss": 0.4819, "mean_token_accuracy": 0.8457817435264587, "num_tokens": 202463765.0, "step": 5304 }, { "epoch": 0.6748505279226561, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.541166305541992, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8569311499595642, "num_tokens": 202505872.0, "step": 5305 }, { "epoch": 0.6749777382012466, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.536158561706543, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.858954131603241, "num_tokens": 202544085.0, "step": 5306 }, { "epoch": 0.6751049484798372, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.536197662353516, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.857276201248169, "num_tokens": 202588237.0, "step": 5307 }, { "epoch": 0.6752321587584277, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.549169540405273, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8535271883010864, "num_tokens": 202628462.0, "step": 5308 }, { "epoch": 0.6753593690370182, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.46638298034668, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8629658222198486, "num_tokens": 202668149.0, "step": 5309 }, { "epoch": 0.6754865793156087, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.557351112365723, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8761979341506958, "num_tokens": 202708651.0, "step": 5310 }, { "epoch": 0.6756137895941993, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.568842887878418, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8652622103691101, "num_tokens": 202738613.0, "step": 5311 }, { "epoch": 0.6757409998727897, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.575892448425293, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.857236385345459, "num_tokens": 202779279.0, "step": 5312 }, { "epoch": 0.6758682101513802, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.528745651245117, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8616855144500732, "num_tokens": 202812095.0, "step": 5313 }, { "epoch": 0.6759954204299707, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.561487197875977, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8563673496246338, "num_tokens": 202855705.0, "step": 5314 }, { "epoch": 0.6761226307085613, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.57032585144043, "learning_rate": 1e-06, "loss": 0.4848, "mean_token_accuracy": 0.8435351252555847, "num_tokens": 202894180.0, "step": 5315 }, { "epoch": 0.6762498409871518, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.583027839660645, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8680144548416138, "num_tokens": 202925842.0, "step": 5316 }, { "epoch": 0.6763770512657423, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.586462020874023, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.866199791431427, "num_tokens": 202959103.0, "step": 5317 }, { "epoch": 0.6765042615443327, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.534912109375, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8603966236114502, "num_tokens": 203000393.0, "step": 5318 }, { "epoch": 0.6766314718229233, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.520984649658203, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8621891736984253, "num_tokens": 203041613.0, "step": 5319 }, { "epoch": 0.6767586821015138, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.54239559173584, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8761792182922363, "num_tokens": 203073744.0, "step": 5320 }, { "epoch": 0.6768858923801043, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.526468276977539, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8462978601455688, "num_tokens": 203108350.0, "step": 5321 }, { "epoch": 0.6770131026586949, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.609623908996582, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8637356162071228, "num_tokens": 203139149.0, "step": 5322 }, { "epoch": 0.6771403129372854, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.525115966796875, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8439416885375977, "num_tokens": 203173953.0, "step": 5323 }, { "epoch": 0.6772675232158758, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.591117858886719, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8707109093666077, "num_tokens": 203208915.0, "step": 5324 }, { "epoch": 0.6773947334944663, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.548318862915039, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8482221364974976, "num_tokens": 203251906.0, "step": 5325 }, { "epoch": 0.6775219437730569, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.571285247802734, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8612645864486694, "num_tokens": 203289460.0, "step": 5326 }, { "epoch": 0.6776491540516474, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.557971954345703, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8690682053565979, "num_tokens": 203325647.0, "step": 5327 }, { "epoch": 0.6777763643302379, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.52109146118164, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8571079969406128, "num_tokens": 203364972.0, "step": 5328 }, { "epoch": 0.6779035746088284, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.546220779418945, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8609413504600525, "num_tokens": 203407778.0, "step": 5329 }, { "epoch": 0.6780307848874189, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.569503784179688, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8611636757850647, "num_tokens": 203443820.0, "step": 5330 }, { "epoch": 0.6781579951660094, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.549819946289062, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8663113713264465, "num_tokens": 203475742.0, "step": 5331 }, { "epoch": 0.6782852054445999, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.614080429077148, "learning_rate": 1e-06, "loss": 0.5126, "mean_token_accuracy": 0.8356066346168518, "num_tokens": 203512617.0, "step": 5332 }, { "epoch": 0.6784124157231904, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.562312126159668, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8460415005683899, "num_tokens": 203546778.0, "step": 5333 }, { "epoch": 0.678539626001781, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.578008651733398, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8623939752578735, "num_tokens": 203585222.0, "step": 5334 }, { "epoch": 0.6786668362803715, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.531670570373535, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8547981977462769, "num_tokens": 203627014.0, "step": 5335 }, { "epoch": 0.6787940465589619, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.57995319366455, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.86671382188797, "num_tokens": 203659189.0, "step": 5336 }, { "epoch": 0.6789212568375524, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.556571960449219, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.872833251953125, "num_tokens": 203695355.0, "step": 5337 }, { "epoch": 0.679048467116143, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.54667854309082, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8690130114555359, "num_tokens": 203733446.0, "step": 5338 }, { "epoch": 0.6791756773947335, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.550576210021973, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8445179462432861, "num_tokens": 203775387.0, "step": 5339 }, { "epoch": 0.679302887673324, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.567240715026855, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8561990261077881, "num_tokens": 203810713.0, "step": 5340 }, { "epoch": 0.6794300979519146, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.570136070251465, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8731000423431396, "num_tokens": 203847144.0, "step": 5341 }, { "epoch": 0.679557308230505, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.505095481872559, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8659441471099854, "num_tokens": 203889716.0, "step": 5342 }, { "epoch": 0.6796845185090955, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.622624397277832, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8470331430435181, "num_tokens": 203927692.0, "step": 5343 }, { "epoch": 0.679811728787686, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.59866714477539, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8479183912277222, "num_tokens": 203960180.0, "step": 5344 }, { "epoch": 0.6799389390662766, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.527363777160645, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.872063398361206, "num_tokens": 203997375.0, "step": 5345 }, { "epoch": 0.6800661493448671, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.571481704711914, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8572131395339966, "num_tokens": 204037588.0, "step": 5346 }, { "epoch": 0.6801933596234576, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.58285903930664, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8536497950553894, "num_tokens": 204080993.0, "step": 5347 }, { "epoch": 0.680320569902048, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.59421157836914, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.862926185131073, "num_tokens": 204116863.0, "step": 5348 }, { "epoch": 0.6804477801806386, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.593489646911621, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8551472425460815, "num_tokens": 204156551.0, "step": 5349 }, { "epoch": 0.6805749904592291, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.575100898742676, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8539233207702637, "num_tokens": 204196210.0, "step": 5350 }, { "epoch": 0.6807022007378196, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.530555725097656, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8587150573730469, "num_tokens": 204234791.0, "step": 5351 }, { "epoch": 0.6808294110164101, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.652762413024902, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8643666505813599, "num_tokens": 204271631.0, "step": 5352 }, { "epoch": 0.6809566212950007, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.572602272033691, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8710825443267822, "num_tokens": 204311879.0, "step": 5353 }, { "epoch": 0.6810838315735911, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.56324291229248, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8641843795776367, "num_tokens": 204357123.0, "step": 5354 }, { "epoch": 0.6812110418521816, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.593122482299805, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8583608865737915, "num_tokens": 204397196.0, "step": 5355 }, { "epoch": 0.6813382521307721, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.609590530395508, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8515306711196899, "num_tokens": 204434654.0, "step": 5356 }, { "epoch": 0.6814654624093627, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.590144157409668, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8834145069122314, "num_tokens": 204466581.0, "step": 5357 }, { "epoch": 0.6815926726879532, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.619199752807617, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8739409446716309, "num_tokens": 204503890.0, "step": 5358 }, { "epoch": 0.6817198829665437, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.568330764770508, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8560737371444702, "num_tokens": 204537160.0, "step": 5359 }, { "epoch": 0.6818470932451343, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.622224807739258, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8716253042221069, "num_tokens": 204570077.0, "step": 5360 }, { "epoch": 0.6819743035237247, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.591878890991211, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8762128353118896, "num_tokens": 204603638.0, "step": 5361 }, { "epoch": 0.6821015138023152, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 10.553764343261719, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8563721179962158, "num_tokens": 204644584.0, "step": 5362 }, { "epoch": 0.6822287240809057, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.613646507263184, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8474375605583191, "num_tokens": 204678459.0, "step": 5363 }, { "epoch": 0.6823559343594963, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.537497520446777, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8581923246383667, "num_tokens": 204712923.0, "step": 5364 }, { "epoch": 0.6824831446380868, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.539336204528809, "learning_rate": 1e-06, "loss": 0.4803, "mean_token_accuracy": 0.8434979915618896, "num_tokens": 204751303.0, "step": 5365 }, { "epoch": 0.6826103549166773, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.617655754089355, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8611293435096741, "num_tokens": 204791685.0, "step": 5366 }, { "epoch": 0.6827375651952677, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.535964965820312, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8619303703308105, "num_tokens": 204827748.0, "step": 5367 }, { "epoch": 0.6828647754738583, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.528081893920898, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8774476647377014, "num_tokens": 204866568.0, "step": 5368 }, { "epoch": 0.6829919857524488, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.631897926330566, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8593786954879761, "num_tokens": 204908334.0, "step": 5369 }, { "epoch": 0.6831191960310393, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.51770305633545, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8636617064476013, "num_tokens": 204949929.0, "step": 5370 }, { "epoch": 0.6832464063096299, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.627172470092773, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.851384699344635, "num_tokens": 204991435.0, "step": 5371 }, { "epoch": 0.6833736165882204, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.582793235778809, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8675636053085327, "num_tokens": 205028434.0, "step": 5372 }, { "epoch": 0.6835008268668108, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.583064079284668, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8551890850067139, "num_tokens": 205064138.0, "step": 5373 }, { "epoch": 0.6836280371454013, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.61384105682373, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8547029495239258, "num_tokens": 205104058.0, "step": 5374 }, { "epoch": 0.6837552474239919, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.588133811950684, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8594352006912231, "num_tokens": 205142609.0, "step": 5375 }, { "epoch": 0.6838824577025824, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.567710876464844, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8709691762924194, "num_tokens": 205180407.0, "step": 5376 }, { "epoch": 0.6840096679811729, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.622551918029785, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8823437690734863, "num_tokens": 205217340.0, "step": 5377 }, { "epoch": 0.6841368782597634, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.598339080810547, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8492530584335327, "num_tokens": 205263671.0, "step": 5378 }, { "epoch": 0.6842640885383539, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.578106880187988, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8582488298416138, "num_tokens": 205302970.0, "step": 5379 }, { "epoch": 0.6843912988169444, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.589005470275879, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8651632070541382, "num_tokens": 205342247.0, "step": 5380 }, { "epoch": 0.6845185090955349, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.575556755065918, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8453316688537598, "num_tokens": 205374897.0, "step": 5381 }, { "epoch": 0.6846457193741254, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.596259117126465, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8583028316497803, "num_tokens": 205411645.0, "step": 5382 }, { "epoch": 0.684772929652716, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.552539825439453, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8605977892875671, "num_tokens": 205450972.0, "step": 5383 }, { "epoch": 0.6849001399313065, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.576465606689453, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8616021871566772, "num_tokens": 205494314.0, "step": 5384 }, { "epoch": 0.6850273502098969, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.637836456298828, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8486628532409668, "num_tokens": 205534696.0, "step": 5385 }, { "epoch": 0.6851545604884874, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.586711883544922, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8729647397994995, "num_tokens": 205575579.0, "step": 5386 }, { "epoch": 0.685281770767078, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.540903091430664, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8643286228179932, "num_tokens": 205609249.0, "step": 5387 }, { "epoch": 0.6854089810456685, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.623164176940918, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8572443127632141, "num_tokens": 205644135.0, "step": 5388 }, { "epoch": 0.685536191324259, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.587323188781738, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8539930582046509, "num_tokens": 205685312.0, "step": 5389 }, { "epoch": 0.6856634016028496, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.571535110473633, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8552238941192627, "num_tokens": 205720525.0, "step": 5390 }, { "epoch": 0.68579061188144, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.56025218963623, "learning_rate": 1e-06, "loss": 0.4924, "mean_token_accuracy": 0.8416929244995117, "num_tokens": 205763358.0, "step": 5391 }, { "epoch": 0.6859178221600305, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.51595687866211, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8645157814025879, "num_tokens": 205801413.0, "step": 5392 }, { "epoch": 0.686045032438621, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.585413932800293, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8620869517326355, "num_tokens": 205840442.0, "step": 5393 }, { "epoch": 0.6861722427172116, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.639324188232422, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8741462230682373, "num_tokens": 205877319.0, "step": 5394 }, { "epoch": 0.6862994529958021, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.547029495239258, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8590962886810303, "num_tokens": 205916186.0, "step": 5395 }, { "epoch": 0.6864266632743926, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.578906059265137, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8503835797309875, "num_tokens": 205956336.0, "step": 5396 }, { "epoch": 0.686553873552983, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.585561752319336, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8596715927124023, "num_tokens": 205989570.0, "step": 5397 }, { "epoch": 0.6866810838315736, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.602960586547852, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8531485795974731, "num_tokens": 206026139.0, "step": 5398 }, { "epoch": 0.6868082941101641, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.514052391052246, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8546191453933716, "num_tokens": 206070059.0, "step": 5399 }, { "epoch": 0.6869355043887546, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.599119186401367, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8641524314880371, "num_tokens": 206104645.0, "step": 5400 }, { "epoch": 0.6870627146673451, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.589760780334473, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8585153818130493, "num_tokens": 206141220.0, "step": 5401 }, { "epoch": 0.6871899249459357, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.53487777709961, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8583466410636902, "num_tokens": 206177917.0, "step": 5402 }, { "epoch": 0.6873171352245261, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.5568265914917, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8595699071884155, "num_tokens": 206213419.0, "step": 5403 }, { "epoch": 0.6874443455031166, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.598434448242188, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8450759649276733, "num_tokens": 206252421.0, "step": 5404 }, { "epoch": 0.6875715557817071, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.577896118164062, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8649711608886719, "num_tokens": 206287195.0, "step": 5405 }, { "epoch": 0.6876987660602977, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.549713134765625, "learning_rate": 1e-06, "loss": 0.5016, "mean_token_accuracy": 0.8412439227104187, "num_tokens": 206328339.0, "step": 5406 }, { "epoch": 0.6878259763388882, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.612850189208984, "learning_rate": 1e-06, "loss": 0.4866, "mean_token_accuracy": 0.8453992605209351, "num_tokens": 206369523.0, "step": 5407 }, { "epoch": 0.6879531866174787, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.542669296264648, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8790810704231262, "num_tokens": 206402864.0, "step": 5408 }, { "epoch": 0.6880803968960693, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.5872163772583, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8490861654281616, "num_tokens": 206439004.0, "step": 5409 }, { "epoch": 0.6882076071746597, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.595768928527832, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8627041578292847, "num_tokens": 206472320.0, "step": 5410 }, { "epoch": 0.6883348174532502, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.551680564880371, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8524122834205627, "num_tokens": 206518649.0, "step": 5411 }, { "epoch": 0.6884620277318407, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.629804611206055, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8747725486755371, "num_tokens": 206551318.0, "step": 5412 }, { "epoch": 0.6885892380104313, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.542801856994629, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8489681482315063, "num_tokens": 206588173.0, "step": 5413 }, { "epoch": 0.6887164482890218, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.614178657531738, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8589783906936646, "num_tokens": 206624013.0, "step": 5414 }, { "epoch": 0.6888436585676123, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.5916109085083, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8676916360855103, "num_tokens": 206668180.0, "step": 5415 }, { "epoch": 0.6889708688462027, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.542201042175293, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8555154204368591, "num_tokens": 206710619.0, "step": 5416 }, { "epoch": 0.6890980791247933, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.604676246643066, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8647761940956116, "num_tokens": 206750121.0, "step": 5417 }, { "epoch": 0.6892252894033838, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.572339057922363, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.87213134765625, "num_tokens": 206789538.0, "step": 5418 }, { "epoch": 0.6893524996819743, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.599267959594727, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8601797819137573, "num_tokens": 206827513.0, "step": 5419 }, { "epoch": 0.6894797099605648, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.641205787658691, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8672386407852173, "num_tokens": 206866563.0, "step": 5420 }, { "epoch": 0.6896069202391554, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.565788269042969, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8705354332923889, "num_tokens": 206906281.0, "step": 5421 }, { "epoch": 0.6897341305177458, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.610712051391602, "learning_rate": 1e-06, "loss": 0.4862, "mean_token_accuracy": 0.8415411114692688, "num_tokens": 206944055.0, "step": 5422 }, { "epoch": 0.6898613407963363, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.572138786315918, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8605889081954956, "num_tokens": 206984421.0, "step": 5423 }, { "epoch": 0.6899885510749268, "ewc_loss": 0.020263671875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 10.589882850646973, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8497614860534668, "num_tokens": 207020319.0, "step": 5424 }, { "epoch": 0.6901157613535174, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.54100227355957, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8636246919631958, "num_tokens": 207061487.0, "step": 5425 }, { "epoch": 0.6902429716321079, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.642855644226074, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8615750670433044, "num_tokens": 207098343.0, "step": 5426 }, { "epoch": 0.6903701819106984, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.526237487792969, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8572183847427368, "num_tokens": 207140381.0, "step": 5427 }, { "epoch": 0.6904973921892888, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.604569435119629, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8736498355865479, "num_tokens": 207173923.0, "step": 5428 }, { "epoch": 0.6906246024678794, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.585238456726074, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8672200441360474, "num_tokens": 207214840.0, "step": 5429 }, { "epoch": 0.6907518127464699, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.609734535217285, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8601033091545105, "num_tokens": 207247850.0, "step": 5430 }, { "epoch": 0.6908790230250604, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.574392318725586, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8611944913864136, "num_tokens": 207286698.0, "step": 5431 }, { "epoch": 0.691006233303651, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.582380294799805, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8552440404891968, "num_tokens": 207322470.0, "step": 5432 }, { "epoch": 0.6911334435822415, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.598653793334961, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8462215662002563, "num_tokens": 207361517.0, "step": 5433 }, { "epoch": 0.6912606538608319, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.625502586364746, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8703000545501709, "num_tokens": 207394504.0, "step": 5434 }, { "epoch": 0.6913878641394224, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.628742218017578, "learning_rate": 1e-06, "loss": 0.4899, "mean_token_accuracy": 0.8418347835540771, "num_tokens": 207430148.0, "step": 5435 }, { "epoch": 0.691515074418013, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.648653984069824, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.857433557510376, "num_tokens": 207461094.0, "step": 5436 }, { "epoch": 0.6916422846966035, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.552983283996582, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8665841221809387, "num_tokens": 207500572.0, "step": 5437 }, { "epoch": 0.691769494975194, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.61882495880127, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8528395891189575, "num_tokens": 207534977.0, "step": 5438 }, { "epoch": 0.6918967052537845, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.603940963745117, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8660871386528015, "num_tokens": 207569809.0, "step": 5439 }, { "epoch": 0.692023915532375, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.553532600402832, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8615148663520813, "num_tokens": 207606785.0, "step": 5440 }, { "epoch": 0.6921511258109655, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.57481861114502, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8588601350784302, "num_tokens": 207648344.0, "step": 5441 }, { "epoch": 0.692278336089556, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.608163833618164, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8677577376365662, "num_tokens": 207680962.0, "step": 5442 }, { "epoch": 0.6924055463681466, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.619734764099121, "learning_rate": 1e-06, "loss": 0.3258, "mean_token_accuracy": 0.892775297164917, "num_tokens": 207709162.0, "step": 5443 }, { "epoch": 0.6925327566467371, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.60200309753418, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8568634986877441, "num_tokens": 207746089.0, "step": 5444 }, { "epoch": 0.6926599669253276, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.61386775970459, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8714685440063477, "num_tokens": 207791235.0, "step": 5445 }, { "epoch": 0.692787177203918, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.629446983337402, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8519705533981323, "num_tokens": 207824527.0, "step": 5446 }, { "epoch": 0.6929143874825086, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.610374450683594, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.870919942855835, "num_tokens": 207863535.0, "step": 5447 }, { "epoch": 0.6930415977610991, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.554372787475586, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8549631834030151, "num_tokens": 207899694.0, "step": 5448 }, { "epoch": 0.6931688080396896, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.62392807006836, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8523787260055542, "num_tokens": 207931440.0, "step": 5449 }, { "epoch": 0.6932960183182801, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.62283706665039, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8590637445449829, "num_tokens": 207963758.0, "step": 5450 }, { "epoch": 0.6934232285968707, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.606243133544922, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.865651547908783, "num_tokens": 208002383.0, "step": 5451 }, { "epoch": 0.6935504388754611, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.583540916442871, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8654540181159973, "num_tokens": 208038723.0, "step": 5452 }, { "epoch": 0.6936776491540516, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.607481002807617, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8656138777732849, "num_tokens": 208071851.0, "step": 5453 }, { "epoch": 0.6938048594326421, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.6237154006958, "learning_rate": 1e-06, "loss": 0.4991, "mean_token_accuracy": 0.8429061770439148, "num_tokens": 208106916.0, "step": 5454 }, { "epoch": 0.6939320697112327, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.55579948425293, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8707603216171265, "num_tokens": 208146883.0, "step": 5455 }, { "epoch": 0.6940592799898232, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.571866989135742, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8634085655212402, "num_tokens": 208187364.0, "step": 5456 }, { "epoch": 0.6941864902684137, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.53879165649414, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.868206262588501, "num_tokens": 208226087.0, "step": 5457 }, { "epoch": 0.6943137005470043, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.572137832641602, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.858110785484314, "num_tokens": 208265193.0, "step": 5458 }, { "epoch": 0.6944409108255947, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.594463348388672, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8745088577270508, "num_tokens": 208305163.0, "step": 5459 }, { "epoch": 0.6945681211041852, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.586054801940918, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8496017456054688, "num_tokens": 208342229.0, "step": 5460 }, { "epoch": 0.6946953313827757, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.575214385986328, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8521324396133423, "num_tokens": 208384606.0, "step": 5461 }, { "epoch": 0.6948225416613663, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.583003044128418, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8799386024475098, "num_tokens": 208416777.0, "step": 5462 }, { "epoch": 0.6949497519399568, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.639405250549316, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8463065028190613, "num_tokens": 208456101.0, "step": 5463 }, { "epoch": 0.6950769622185473, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.62405776977539, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8708925247192383, "num_tokens": 208489201.0, "step": 5464 }, { "epoch": 0.6952041724971377, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.591202735900879, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8749390840530396, "num_tokens": 208525006.0, "step": 5465 }, { "epoch": 0.6953313827757283, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.5987548828125, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.858436107635498, "num_tokens": 208568997.0, "step": 5466 }, { "epoch": 0.6954585930543188, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.584930419921875, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8531948328018188, "num_tokens": 208605925.0, "step": 5467 }, { "epoch": 0.6955858033329093, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.638680458068848, "learning_rate": 1e-06, "loss": 0.52, "mean_token_accuracy": 0.8366572260856628, "num_tokens": 208646290.0, "step": 5468 }, { "epoch": 0.6957130136114998, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.554287910461426, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8626670837402344, "num_tokens": 208686424.0, "step": 5469 }, { "epoch": 0.6958402238900904, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.610033988952637, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8630832433700562, "num_tokens": 208719643.0, "step": 5470 }, { "epoch": 0.6959674341686808, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.574044227600098, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.845270574092865, "num_tokens": 208760804.0, "step": 5471 }, { "epoch": 0.6960946444472713, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.586647033691406, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8555570244789124, "num_tokens": 208803022.0, "step": 5472 }, { "epoch": 0.6962218547258618, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.563909530639648, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8617469072341919, "num_tokens": 208844205.0, "step": 5473 }, { "epoch": 0.6963490650044524, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.543261528015137, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8716011643409729, "num_tokens": 208886882.0, "step": 5474 }, { "epoch": 0.6964762752830429, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.663187980651855, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8577772378921509, "num_tokens": 208926649.0, "step": 5475 }, { "epoch": 0.6966034855616334, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.578141212463379, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.854623019695282, "num_tokens": 208963403.0, "step": 5476 }, { "epoch": 0.6967306958402238, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.650050163269043, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8535503149032593, "num_tokens": 209002475.0, "step": 5477 }, { "epoch": 0.6968579061188144, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.611241340637207, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8607497811317444, "num_tokens": 209044216.0, "step": 5478 }, { "epoch": 0.6969851163974049, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.635639190673828, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8506540656089783, "num_tokens": 209080774.0, "step": 5479 }, { "epoch": 0.6971123266759954, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.609885215759277, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8664460778236389, "num_tokens": 209117167.0, "step": 5480 }, { "epoch": 0.697239536954586, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.69915771484375, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8556585907936096, "num_tokens": 209148872.0, "step": 5481 }, { "epoch": 0.6973667472331765, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.631155014038086, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8622248768806458, "num_tokens": 209187912.0, "step": 5482 }, { "epoch": 0.6974939575117669, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.578825950622559, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8673725128173828, "num_tokens": 209227055.0, "step": 5483 }, { "epoch": 0.6976211677903574, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.628522872924805, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8747152090072632, "num_tokens": 209255343.0, "step": 5484 }, { "epoch": 0.697748378068948, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.6244478225708, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8543658256530762, "num_tokens": 209294621.0, "step": 5485 }, { "epoch": 0.6978755883475385, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.604883193969727, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8763372898101807, "num_tokens": 209336519.0, "step": 5486 }, { "epoch": 0.698002798626129, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.622394561767578, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8582932949066162, "num_tokens": 209372913.0, "step": 5487 }, { "epoch": 0.6981300089047195, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.564756393432617, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8527126908302307, "num_tokens": 209417768.0, "step": 5488 }, { "epoch": 0.69825721918331, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.613347053527832, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8636729121208191, "num_tokens": 209451221.0, "step": 5489 }, { "epoch": 0.6983844294619005, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.660715103149414, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8597254157066345, "num_tokens": 209488710.0, "step": 5490 }, { "epoch": 0.698511639740491, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.571520805358887, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8576840758323669, "num_tokens": 209525166.0, "step": 5491 }, { "epoch": 0.6986388500190815, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.574325561523438, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.863499104976654, "num_tokens": 209564804.0, "step": 5492 }, { "epoch": 0.6987660602976721, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.606684684753418, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8646460175514221, "num_tokens": 209607155.0, "step": 5493 }, { "epoch": 0.6988932705762626, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.624159812927246, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8512436151504517, "num_tokens": 209641226.0, "step": 5494 }, { "epoch": 0.699020480854853, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.570131301879883, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.858227014541626, "num_tokens": 209679180.0, "step": 5495 }, { "epoch": 0.6991476911334435, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.615930557250977, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8520088195800781, "num_tokens": 209716412.0, "step": 5496 }, { "epoch": 0.6992749014120341, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.589010238647461, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.851799488067627, "num_tokens": 209755697.0, "step": 5497 }, { "epoch": 0.6994021116906246, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.652382850646973, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8681261539459229, "num_tokens": 209793654.0, "step": 5498 }, { "epoch": 0.6995293219692151, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.579270362854004, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8589221239089966, "num_tokens": 209830703.0, "step": 5499 }, { "epoch": 0.6996565322478057, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.651060104370117, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8607229590415955, "num_tokens": 209868994.0, "step": 5500 }, { "epoch": 0.6997837425263961, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.56502628326416, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8708956241607666, "num_tokens": 209909605.0, "step": 5501 }, { "epoch": 0.6999109528049866, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.64134407043457, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8672594428062439, "num_tokens": 209945196.0, "step": 5502 }, { "epoch": 0.7000381630835771, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.610093116760254, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8609645962715149, "num_tokens": 209988585.0, "step": 5503 }, { "epoch": 0.7001653733621677, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.640856742858887, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8606553077697754, "num_tokens": 210032084.0, "step": 5504 }, { "epoch": 0.7002925836407582, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.56932258605957, "learning_rate": 1e-06, "loss": 0.4737, "mean_token_accuracy": 0.8504548072814941, "num_tokens": 210070639.0, "step": 5505 }, { "epoch": 0.7004197939193487, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.657906532287598, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8691624402999878, "num_tokens": 210113793.0, "step": 5506 }, { "epoch": 0.7005470041979391, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.586195945739746, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8529781699180603, "num_tokens": 210153517.0, "step": 5507 }, { "epoch": 0.7006742144765297, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.639870643615723, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8563810586929321, "num_tokens": 210186818.0, "step": 5508 }, { "epoch": 0.7008014247551202, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.611241340637207, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8528734445571899, "num_tokens": 210224283.0, "step": 5509 }, { "epoch": 0.7009286350337107, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.591901779174805, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8555730581283569, "num_tokens": 210263439.0, "step": 5510 }, { "epoch": 0.7010558453123013, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.569320678710938, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8624988794326782, "num_tokens": 210305887.0, "step": 5511 }, { "epoch": 0.7011830555908918, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.603692054748535, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8601043224334717, "num_tokens": 210340118.0, "step": 5512 }, { "epoch": 0.7013102658694823, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.628338813781738, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8613930344581604, "num_tokens": 210376098.0, "step": 5513 }, { "epoch": 0.7014374761480727, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.558911323547363, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8651982545852661, "num_tokens": 210418801.0, "step": 5514 }, { "epoch": 0.7015646864266633, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.619847297668457, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8511781096458435, "num_tokens": 210452392.0, "step": 5515 }, { "epoch": 0.7016918967052538, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.603582382202148, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8676308393478394, "num_tokens": 210489697.0, "step": 5516 }, { "epoch": 0.7018191069838443, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.621460914611816, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.845477819442749, "num_tokens": 210530212.0, "step": 5517 }, { "epoch": 0.7019463172624348, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.672853469848633, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8668678998947144, "num_tokens": 210564974.0, "step": 5518 }, { "epoch": 0.7020735275410254, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.599002838134766, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8615531921386719, "num_tokens": 210601272.0, "step": 5519 }, { "epoch": 0.7022007378196158, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.653071403503418, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8638176918029785, "num_tokens": 210636908.0, "step": 5520 }, { "epoch": 0.7023279480982063, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.637495040893555, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.857887864112854, "num_tokens": 210672629.0, "step": 5521 }, { "epoch": 0.7024551583767968, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.610407829284668, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8752345442771912, "num_tokens": 210703531.0, "step": 5522 }, { "epoch": 0.7025823686553874, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.589628219604492, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8736670017242432, "num_tokens": 210742946.0, "step": 5523 }, { "epoch": 0.7027095789339779, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.639068603515625, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8696109056472778, "num_tokens": 210778345.0, "step": 5524 }, { "epoch": 0.7028367892125684, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.59682559967041, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8667106628417969, "num_tokens": 210821544.0, "step": 5525 }, { "epoch": 0.7029639994911588, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.628934860229492, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8647490739822388, "num_tokens": 210858060.0, "step": 5526 }, { "epoch": 0.7030912097697494, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.59785270690918, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8553948402404785, "num_tokens": 210900365.0, "step": 5527 }, { "epoch": 0.7032184200483399, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.683074951171875, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.855866551399231, "num_tokens": 210941118.0, "step": 5528 }, { "epoch": 0.7033456303269304, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.564029693603516, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8678924441337585, "num_tokens": 210979865.0, "step": 5529 }, { "epoch": 0.703472840605521, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.61616325378418, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8613778352737427, "num_tokens": 211016385.0, "step": 5530 }, { "epoch": 0.7036000508841115, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.64438533782959, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8494650721549988, "num_tokens": 211059071.0, "step": 5531 }, { "epoch": 0.7037272611627019, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.608717918395996, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8638328313827515, "num_tokens": 211094189.0, "step": 5532 }, { "epoch": 0.7038544714412924, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.606834411621094, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8757940530776978, "num_tokens": 211130524.0, "step": 5533 }, { "epoch": 0.703981681719883, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.627495765686035, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8545291423797607, "num_tokens": 211172335.0, "step": 5534 }, { "epoch": 0.7041088919984735, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.600170135498047, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8628344535827637, "num_tokens": 211205406.0, "step": 5535 }, { "epoch": 0.704236102277064, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.615227699279785, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8574404716491699, "num_tokens": 211246443.0, "step": 5536 }, { "epoch": 0.7043633125556545, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.670196533203125, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8756146430969238, "num_tokens": 211286707.0, "step": 5537 }, { "epoch": 0.704490522834245, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.578125, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.872348427772522, "num_tokens": 211319448.0, "step": 5538 }, { "epoch": 0.7046177331128355, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.595563888549805, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8558616042137146, "num_tokens": 211363747.0, "step": 5539 }, { "epoch": 0.704744943391426, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.622964859008789, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8678653240203857, "num_tokens": 211408080.0, "step": 5540 }, { "epoch": 0.7048721536700165, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.606171607971191, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8509785532951355, "num_tokens": 211445656.0, "step": 5541 }, { "epoch": 0.7049993639486071, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.631021499633789, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.870186448097229, "num_tokens": 211484342.0, "step": 5542 }, { "epoch": 0.7051265742271976, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.643828392028809, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8567484617233276, "num_tokens": 211518535.0, "step": 5543 }, { "epoch": 0.705253784505788, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.651174545288086, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8548247814178467, "num_tokens": 211557236.0, "step": 5544 }, { "epoch": 0.7053809947843785, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.616724014282227, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8727997541427612, "num_tokens": 211599216.0, "step": 5545 }, { "epoch": 0.7055082050629691, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.637516975402832, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8433310389518738, "num_tokens": 211638930.0, "step": 5546 }, { "epoch": 0.7056354153415596, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.685938835144043, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8535335063934326, "num_tokens": 211680765.0, "step": 5547 }, { "epoch": 0.7057626256201501, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.569108963012695, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8747622966766357, "num_tokens": 211719037.0, "step": 5548 }, { "epoch": 0.7058898358987407, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.675300598144531, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8555821180343628, "num_tokens": 211759655.0, "step": 5549 }, { "epoch": 0.7060170461773311, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.603273391723633, "learning_rate": 1e-06, "loss": 0.492, "mean_token_accuracy": 0.8441521525382996, "num_tokens": 211800405.0, "step": 5550 }, { "epoch": 0.7061442564559216, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.580160140991211, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.84331214427948, "num_tokens": 211846057.0, "step": 5551 }, { "epoch": 0.7062714667345121, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.6565580368042, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8578891158103943, "num_tokens": 211885944.0, "step": 5552 }, { "epoch": 0.7063986770131027, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.631902694702148, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8644288778305054, "num_tokens": 211928727.0, "step": 5553 }, { "epoch": 0.7065258872916932, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.622386932373047, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8509981036186218, "num_tokens": 211967593.0, "step": 5554 }, { "epoch": 0.7066530975702837, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.655104637145996, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8565240502357483, "num_tokens": 212001023.0, "step": 5555 }, { "epoch": 0.7067803078488741, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.621085166931152, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8523130416870117, "num_tokens": 212045611.0, "step": 5556 }, { "epoch": 0.7069075181274647, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.669119834899902, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8561407327651978, "num_tokens": 212087167.0, "step": 5557 }, { "epoch": 0.7070347284060552, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.635308265686035, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8618575930595398, "num_tokens": 212124254.0, "step": 5558 }, { "epoch": 0.7071619386846457, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.663832664489746, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8546267747879028, "num_tokens": 212162706.0, "step": 5559 }, { "epoch": 0.7072891489632362, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 10.659602165222168, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8605120778083801, "num_tokens": 212197359.0, "step": 5560 }, { "epoch": 0.7074163592418268, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.65955638885498, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8670933246612549, "num_tokens": 212237300.0, "step": 5561 }, { "epoch": 0.7075435695204173, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.61498737335205, "learning_rate": 1e-06, "loss": 0.4907, "mean_token_accuracy": 0.8412874937057495, "num_tokens": 212283472.0, "step": 5562 }, { "epoch": 0.7076707797990077, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.660722732543945, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8551455736160278, "num_tokens": 212321433.0, "step": 5563 }, { "epoch": 0.7077979900775982, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.59619426727295, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8581230640411377, "num_tokens": 212362234.0, "step": 5564 }, { "epoch": 0.7079252003561888, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.711034774780273, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8513141870498657, "num_tokens": 212399259.0, "step": 5565 }, { "epoch": 0.7080524106347793, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.600613594055176, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.860711932182312, "num_tokens": 212438489.0, "step": 5566 }, { "epoch": 0.7081796209133698, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.647441864013672, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8521933555603027, "num_tokens": 212482283.0, "step": 5567 }, { "epoch": 0.7083068311919604, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.603921890258789, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8552303910255432, "num_tokens": 212522749.0, "step": 5568 }, { "epoch": 0.7084340414705508, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.676533699035645, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8516435027122498, "num_tokens": 212559937.0, "step": 5569 }, { "epoch": 0.7085612517491413, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.624893188476562, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8654605150222778, "num_tokens": 212595578.0, "step": 5570 }, { "epoch": 0.7086884620277318, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.630788803100586, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8717377185821533, "num_tokens": 212634085.0, "step": 5571 }, { "epoch": 0.7088156723063224, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.628839492797852, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8497506976127625, "num_tokens": 212673557.0, "step": 5572 }, { "epoch": 0.7089428825849129, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.644966125488281, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8693784475326538, "num_tokens": 212711142.0, "step": 5573 }, { "epoch": 0.7090700928635034, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.618281364440918, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8609282970428467, "num_tokens": 212745570.0, "step": 5574 }, { "epoch": 0.7091973031420938, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.600509643554688, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8641340732574463, "num_tokens": 212783048.0, "step": 5575 }, { "epoch": 0.7093245134206844, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.628296852111816, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8773254156112671, "num_tokens": 212821231.0, "step": 5576 }, { "epoch": 0.7094517236992749, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.604544639587402, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8584563732147217, "num_tokens": 212861715.0, "step": 5577 }, { "epoch": 0.7095789339778654, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.617077827453613, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.849585771560669, "num_tokens": 212900655.0, "step": 5578 }, { "epoch": 0.709706144256456, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.654911041259766, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8512683510780334, "num_tokens": 212939308.0, "step": 5579 }, { "epoch": 0.7098333545350465, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.594013214111328, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8636152744293213, "num_tokens": 212981174.0, "step": 5580 }, { "epoch": 0.7099605648136369, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.63631534576416, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8698945641517639, "num_tokens": 213016401.0, "step": 5581 }, { "epoch": 0.7100877750922274, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.619291305541992, "learning_rate": 1e-06, "loss": 0.4781, "mean_token_accuracy": 0.8478774428367615, "num_tokens": 213058480.0, "step": 5582 }, { "epoch": 0.710214985370818, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.598532676696777, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8722686767578125, "num_tokens": 213093527.0, "step": 5583 }, { "epoch": 0.7103421956494085, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.683451652526855, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8569604158401489, "num_tokens": 213126778.0, "step": 5584 }, { "epoch": 0.710469405927999, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.620394706726074, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8632769584655762, "num_tokens": 213170863.0, "step": 5585 }, { "epoch": 0.7105966162065895, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.618666648864746, "learning_rate": 1e-06, "loss": 0.4998, "mean_token_accuracy": 0.8357141017913818, "num_tokens": 213211336.0, "step": 5586 }, { "epoch": 0.71072382648518, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.641202926635742, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8668670058250427, "num_tokens": 213245262.0, "step": 5587 }, { "epoch": 0.7108510367637705, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.718677520751953, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8517007827758789, "num_tokens": 213282150.0, "step": 5588 }, { "epoch": 0.710978247042361, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.631156921386719, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8655194640159607, "num_tokens": 213325573.0, "step": 5589 }, { "epoch": 0.7111054573209515, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.682417869567871, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8681814074516296, "num_tokens": 213363738.0, "step": 5590 }, { "epoch": 0.7112326675995421, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.679558753967285, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8652819395065308, "num_tokens": 213404796.0, "step": 5591 }, { "epoch": 0.7113598778781326, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.633620262145996, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8520364761352539, "num_tokens": 213436711.0, "step": 5592 }, { "epoch": 0.711487088156723, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.692403793334961, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8625841736793518, "num_tokens": 213473458.0, "step": 5593 }, { "epoch": 0.7116142984353135, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.686640739440918, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8708643913269043, "num_tokens": 213507355.0, "step": 5594 }, { "epoch": 0.7117415087139041, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.658966064453125, "learning_rate": 1e-06, "loss": 0.4884, "mean_token_accuracy": 0.8428100347518921, "num_tokens": 213543771.0, "step": 5595 }, { "epoch": 0.7118687189924946, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.701892852783203, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8585739135742188, "num_tokens": 213578317.0, "step": 5596 }, { "epoch": 0.7119959292710851, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.64350700378418, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.856787383556366, "num_tokens": 213615795.0, "step": 5597 }, { "epoch": 0.7121231395496757, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.683614730834961, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8689470291137695, "num_tokens": 213655452.0, "step": 5598 }, { "epoch": 0.7122503498282661, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.675592422485352, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8516976833343506, "num_tokens": 213688079.0, "step": 5599 }, { "epoch": 0.7123775601068566, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.694106101989746, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.877037763595581, "num_tokens": 213725310.0, "step": 5600 }, { "epoch": 0.7125047703854471, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.603633880615234, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8541902303695679, "num_tokens": 213757950.0, "step": 5601 }, { "epoch": 0.7126319806640377, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.702825546264648, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8482123613357544, "num_tokens": 213794146.0, "step": 5602 }, { "epoch": 0.7127591909426282, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.647590637207031, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8638283610343933, "num_tokens": 213834922.0, "step": 5603 }, { "epoch": 0.7128864012212187, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.625367164611816, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8785954713821411, "num_tokens": 213879651.0, "step": 5604 }, { "epoch": 0.7130136114998091, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.665645599365234, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.859417200088501, "num_tokens": 213921833.0, "step": 5605 }, { "epoch": 0.7131408217783997, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.607808113098145, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8648660778999329, "num_tokens": 213962909.0, "step": 5606 }, { "epoch": 0.7132680320569902, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.695225715637207, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8626251220703125, "num_tokens": 214002030.0, "step": 5607 }, { "epoch": 0.7133952423355807, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.664257049560547, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8542249202728271, "num_tokens": 214037094.0, "step": 5608 }, { "epoch": 0.7135224526141712, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.634180068969727, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8643355369567871, "num_tokens": 214076649.0, "step": 5609 }, { "epoch": 0.7136496628927618, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.645829200744629, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8566986322402954, "num_tokens": 214114194.0, "step": 5610 }, { "epoch": 0.7137768731713523, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.695241928100586, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.86216801404953, "num_tokens": 214149652.0, "step": 5611 }, { "epoch": 0.7139040834499427, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.656002044677734, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8698657751083374, "num_tokens": 214189068.0, "step": 5612 }, { "epoch": 0.7140312937285332, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.672222137451172, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8487764596939087, "num_tokens": 214220251.0, "step": 5613 }, { "epoch": 0.7141585040071238, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.738912582397461, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8580601811408997, "num_tokens": 214256705.0, "step": 5614 }, { "epoch": 0.7142857142857143, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.638001441955566, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8744151592254639, "num_tokens": 214292563.0, "step": 5615 }, { "epoch": 0.7144129245643048, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.635087013244629, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8543192148208618, "num_tokens": 214333843.0, "step": 5616 }, { "epoch": 0.7145401348428954, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.667081832885742, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8716391324996948, "num_tokens": 214372272.0, "step": 5617 }, { "epoch": 0.7146673451214858, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.632207870483398, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8670505285263062, "num_tokens": 214416364.0, "step": 5618 }, { "epoch": 0.7147945554000763, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.659834861755371, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8582388758659363, "num_tokens": 214453102.0, "step": 5619 }, { "epoch": 0.7149217656786668, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.694927215576172, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8648198246955872, "num_tokens": 214492374.0, "step": 5620 }, { "epoch": 0.7150489759572574, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.69476318359375, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8609901070594788, "num_tokens": 214530918.0, "step": 5621 }, { "epoch": 0.7151761862358479, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.62876033782959, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8758330345153809, "num_tokens": 214570227.0, "step": 5622 }, { "epoch": 0.7153033965144384, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.688610076904297, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8640629053115845, "num_tokens": 214609519.0, "step": 5623 }, { "epoch": 0.7154306067930288, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.640385627746582, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8578923940658569, "num_tokens": 214643564.0, "step": 5624 }, { "epoch": 0.7155578170716194, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.714070320129395, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8690589666366577, "num_tokens": 214677632.0, "step": 5625 }, { "epoch": 0.7156850273502099, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.629342079162598, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8638029098510742, "num_tokens": 214710363.0, "step": 5626 }, { "epoch": 0.7158122376288004, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.661497116088867, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8679178953170776, "num_tokens": 214746982.0, "step": 5627 }, { "epoch": 0.715939447907391, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.696417808532715, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8515291213989258, "num_tokens": 214786447.0, "step": 5628 }, { "epoch": 0.7160666581859815, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.663422584533691, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8694756031036377, "num_tokens": 214824627.0, "step": 5629 }, { "epoch": 0.7161938684645719, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.678227424621582, "learning_rate": 1e-06, "loss": 0.4976, "mean_token_accuracy": 0.8422518968582153, "num_tokens": 214863051.0, "step": 5630 }, { "epoch": 0.7163210787431624, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.696596145629883, "learning_rate": 1e-06, "loss": 0.4933, "mean_token_accuracy": 0.8479457497596741, "num_tokens": 214900524.0, "step": 5631 }, { "epoch": 0.716448289021753, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.699036598205566, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8667349815368652, "num_tokens": 214943949.0, "step": 5632 }, { "epoch": 0.7165754993003435, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.632135391235352, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8659812211990356, "num_tokens": 214979126.0, "step": 5633 }, { "epoch": 0.716702709578934, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.69106674194336, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8610338568687439, "num_tokens": 215017740.0, "step": 5634 }, { "epoch": 0.7168299198575245, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.649114608764648, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8693290948867798, "num_tokens": 215054472.0, "step": 5635 }, { "epoch": 0.716957130136115, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.691352844238281, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8637285232543945, "num_tokens": 215086894.0, "step": 5636 }, { "epoch": 0.7170843404147055, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.696834564208984, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8624075651168823, "num_tokens": 215128157.0, "step": 5637 }, { "epoch": 0.717211550693296, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.64852523803711, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8683949708938599, "num_tokens": 215161008.0, "step": 5638 }, { "epoch": 0.7173387609718865, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.67646598815918, "learning_rate": 1e-06, "loss": 0.5069, "mean_token_accuracy": 0.8444187641143799, "num_tokens": 215197157.0, "step": 5639 }, { "epoch": 0.7174659712504771, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.637768745422363, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.868504524230957, "num_tokens": 215240270.0, "step": 5640 }, { "epoch": 0.7175931815290676, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.633358001708984, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8687065839767456, "num_tokens": 215279828.0, "step": 5641 }, { "epoch": 0.717720391807658, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.689437866210938, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8560185432434082, "num_tokens": 215315271.0, "step": 5642 }, { "epoch": 0.7178476020862485, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.688074111938477, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8630287051200867, "num_tokens": 215352621.0, "step": 5643 }, { "epoch": 0.7179748123648391, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.631736755371094, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8638925552368164, "num_tokens": 215390231.0, "step": 5644 }, { "epoch": 0.7181020226434296, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.657957077026367, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8753368854522705, "num_tokens": 215426812.0, "step": 5645 }, { "epoch": 0.7182292329220201, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.636674880981445, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8503885269165039, "num_tokens": 215467551.0, "step": 5646 }, { "epoch": 0.7183564432006107, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.691060066223145, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.869999885559082, "num_tokens": 215508559.0, "step": 5647 }, { "epoch": 0.7184836534792011, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.62992000579834, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8544399738311768, "num_tokens": 215548709.0, "step": 5648 }, { "epoch": 0.7186108637577916, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.642492294311523, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8622454404830933, "num_tokens": 215593844.0, "step": 5649 }, { "epoch": 0.7187380740363821, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.67197322845459, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8671814203262329, "num_tokens": 215633853.0, "step": 5650 }, { "epoch": 0.7188652843149727, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.720711708068848, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8601236343383789, "num_tokens": 215667272.0, "step": 5651 }, { "epoch": 0.7189924945935632, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.677876472473145, "learning_rate": 1e-06, "loss": 0.4823, "mean_token_accuracy": 0.8445335030555725, "num_tokens": 215705889.0, "step": 5652 }, { "epoch": 0.7191197048721537, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.68056583404541, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8706803917884827, "num_tokens": 215749143.0, "step": 5653 }, { "epoch": 0.7192469151507441, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.67334270477295, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8561303615570068, "num_tokens": 215787583.0, "step": 5654 }, { "epoch": 0.7193741254293347, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.675175666809082, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8619344830513, "num_tokens": 215825552.0, "step": 5655 }, { "epoch": 0.7195013357079252, "ewc_loss": 0.0205078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.663753509521484, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8706647157669067, "num_tokens": 215858310.0, "step": 5656 }, { "epoch": 0.7196285459865157, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.763193130493164, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8503322601318359, "num_tokens": 215898403.0, "step": 5657 }, { "epoch": 0.7197557562651062, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.64734935760498, "learning_rate": 1e-06, "loss": 0.5232, "mean_token_accuracy": 0.8386799097061157, "num_tokens": 215935930.0, "step": 5658 }, { "epoch": 0.7198829665436968, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.729998588562012, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8586443662643433, "num_tokens": 215971509.0, "step": 5659 }, { "epoch": 0.7200101768222873, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.684736251831055, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8671212792396545, "num_tokens": 216008696.0, "step": 5660 }, { "epoch": 0.7201373871008777, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.69244384765625, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8615697622299194, "num_tokens": 216045861.0, "step": 5661 }, { "epoch": 0.7202645973794682, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.687163352966309, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8710228204727173, "num_tokens": 216081307.0, "step": 5662 }, { "epoch": 0.7203918076580588, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.635608673095703, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8711714744567871, "num_tokens": 216118506.0, "step": 5663 }, { "epoch": 0.7205190179366493, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.653173446655273, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8544213175773621, "num_tokens": 216159000.0, "step": 5664 }, { "epoch": 0.7206462282152398, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.639618873596191, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8544019460678101, "num_tokens": 216204712.0, "step": 5665 }, { "epoch": 0.7207734384938304, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.658623695373535, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8543047904968262, "num_tokens": 216235001.0, "step": 5666 }, { "epoch": 0.7209006487724208, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.671587944030762, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8554920554161072, "num_tokens": 216271383.0, "step": 5667 }, { "epoch": 0.7210278590510113, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.709029197692871, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8624762892723083, "num_tokens": 216307923.0, "step": 5668 }, { "epoch": 0.7211550693296018, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.618724822998047, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8807211518287659, "num_tokens": 216345147.0, "step": 5669 }, { "epoch": 0.7212822796081924, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.655913352966309, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8541895151138306, "num_tokens": 216386991.0, "step": 5670 }, { "epoch": 0.7214094898867829, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.682646751403809, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.864966094493866, "num_tokens": 216425534.0, "step": 5671 }, { "epoch": 0.7215367001653734, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.703583717346191, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8613569140434265, "num_tokens": 216466686.0, "step": 5672 }, { "epoch": 0.7216639104439638, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.663994789123535, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8729473948478699, "num_tokens": 216505245.0, "step": 5673 }, { "epoch": 0.7217911207225544, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.793439865112305, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8576897382736206, "num_tokens": 216538024.0, "step": 5674 }, { "epoch": 0.7219183310011449, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.651936531066895, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8609937429428101, "num_tokens": 216573722.0, "step": 5675 }, { "epoch": 0.7220455412797354, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.680785179138184, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8682406544685364, "num_tokens": 216610362.0, "step": 5676 }, { "epoch": 0.7221727515583259, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.746380805969238, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8535057902336121, "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": 10.636382102966309, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8699667453765869, "num_tokens": 216680324.0, "step": 5678 }, { "epoch": 0.7224271721155069, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.657289505004883, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8652210235595703, "num_tokens": 216712773.0, "step": 5679 }, { "epoch": 0.7225543823940974, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.719284057617188, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8537743091583252, "num_tokens": 216752538.0, "step": 5680 }, { "epoch": 0.722681592672688, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.615859985351562, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8640013933181763, "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": 10.611187934875488, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8587063550949097, "num_tokens": 216831729.0, "step": 5682 }, { "epoch": 0.722936013229869, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.643961906433105, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8627253770828247, "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": 10.670631408691406, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8440912961959839, "num_tokens": 216906672.0, "step": 5684 }, { "epoch": 0.72319043378705, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.617646217346191, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8684133291244507, "num_tokens": 216941587.0, "step": 5685 }, { "epoch": 0.7233176440656405, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.63206958770752, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8544975519180298, "num_tokens": 216981812.0, "step": 5686 }, { "epoch": 0.723444854344231, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.568828582763672, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8708211183547974, "num_tokens": 217024137.0, "step": 5687 }, { "epoch": 0.7235720646228215, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.631709098815918, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8738579750061035, "num_tokens": 217063753.0, "step": 5688 }, { "epoch": 0.7236992749014121, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.663747787475586, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8601888418197632, "num_tokens": 217105611.0, "step": 5689 }, { "epoch": 0.7238264851800026, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.6690673828125, "learning_rate": 1e-06, "loss": 0.4859, "mean_token_accuracy": 0.841910719871521, "num_tokens": 217145989.0, "step": 5690 }, { "epoch": 0.723953695458593, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.58797550201416, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8793519735336304, "num_tokens": 217183726.0, "step": 5691 }, { "epoch": 0.7240809057371835, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.688488960266113, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.859477698802948, "num_tokens": 217226644.0, "step": 5692 }, { "epoch": 0.7242081160157741, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.684906959533691, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8477579951286316, "num_tokens": 217263327.0, "step": 5693 }, { "epoch": 0.7243353262943646, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.70431900024414, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8663185834884644, "num_tokens": 217300595.0, "step": 5694 }, { "epoch": 0.7244625365729551, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.693926811218262, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8477469682693481, "num_tokens": 217336385.0, "step": 5695 }, { "epoch": 0.7245897468515456, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.692689895629883, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8680477142333984, "num_tokens": 217374309.0, "step": 5696 }, { "epoch": 0.7247169571301361, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.673310279846191, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8525444865226746, "num_tokens": 217408386.0, "step": 5697 }, { "epoch": 0.7248441674087266, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.693807601928711, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8481143116950989, "num_tokens": 217442855.0, "step": 5698 }, { "epoch": 0.7249713776873171, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.669933319091797, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8662185072898865, "num_tokens": 217476892.0, "step": 5699 }, { "epoch": 0.7250985879659076, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.665996551513672, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8592796325683594, "num_tokens": 217513569.0, "step": 5700 }, { "epoch": 0.7252257982444982, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.73972225189209, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8559905290603638, "num_tokens": 217545585.0, "step": 5701 }, { "epoch": 0.7253530085230887, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.64681339263916, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8643120527267456, "num_tokens": 217583156.0, "step": 5702 }, { "epoch": 0.7254802188016791, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.663949012756348, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8619700074195862, "num_tokens": 217619630.0, "step": 5703 }, { "epoch": 0.7256074290802697, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.694568634033203, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8723744750022888, "num_tokens": 217653616.0, "step": 5704 }, { "epoch": 0.7257346393588602, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.663723945617676, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8570383191108704, "num_tokens": 217692758.0, "step": 5705 }, { "epoch": 0.7258618496374507, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.663260459899902, "learning_rate": 1e-06, "loss": 0.4831, "mean_token_accuracy": 0.8467284440994263, "num_tokens": 217734175.0, "step": 5706 }, { "epoch": 0.7259890599160412, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.701038360595703, "learning_rate": 1e-06, "loss": 0.4949, "mean_token_accuracy": 0.8392165899276733, "num_tokens": 217773372.0, "step": 5707 }, { "epoch": 0.7261162701946318, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.660805702209473, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.871759295463562, "num_tokens": 217809174.0, "step": 5708 }, { "epoch": 0.7262434804732223, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.70693302154541, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8543657064437866, "num_tokens": 217845882.0, "step": 5709 }, { "epoch": 0.7263706907518127, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.668259620666504, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8569063544273376, "num_tokens": 217885798.0, "step": 5710 }, { "epoch": 0.7264979010304032, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.703621864318848, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8574270009994507, "num_tokens": 217924967.0, "step": 5711 }, { "epoch": 0.7266251113089938, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.651656150817871, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.872429370880127, "num_tokens": 217967381.0, "step": 5712 }, { "epoch": 0.7267523215875843, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.675905227661133, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8673981428146362, "num_tokens": 218001097.0, "step": 5713 }, { "epoch": 0.7268795318661748, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.681553840637207, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8633981943130493, "num_tokens": 218040705.0, "step": 5714 }, { "epoch": 0.7270067421447653, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.702795028686523, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8613153696060181, "num_tokens": 218077549.0, "step": 5715 }, { "epoch": 0.7271339524233558, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.646733283996582, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8602554202079773, "num_tokens": 218116251.0, "step": 5716 }, { "epoch": 0.7272611627019463, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.689580917358398, "learning_rate": 1e-06, "loss": 0.5092, "mean_token_accuracy": 0.8346518874168396, "num_tokens": 218155798.0, "step": 5717 }, { "epoch": 0.7273883729805368, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.686771392822266, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8665978908538818, "num_tokens": 218191277.0, "step": 5718 }, { "epoch": 0.7275155832591274, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.718456268310547, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8670703172683716, "num_tokens": 218233165.0, "step": 5719 }, { "epoch": 0.7276427935377179, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 10.667330741882324, "learning_rate": 1e-06, "loss": 0.4957, "mean_token_accuracy": 0.8412971496582031, "num_tokens": 218271263.0, "step": 5720 }, { "epoch": 0.7277700038163084, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.712008476257324, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8634163737297058, "num_tokens": 218309081.0, "step": 5721 }, { "epoch": 0.7278972140948988, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.692599296569824, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8568587303161621, "num_tokens": 218348314.0, "step": 5722 }, { "epoch": 0.7280244243734894, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.607959747314453, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.864894449710846, "num_tokens": 218387556.0, "step": 5723 }, { "epoch": 0.7281516346520799, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.74532413482666, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8531425595283508, "num_tokens": 218423130.0, "step": 5724 }, { "epoch": 0.7282788449306704, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.667901039123535, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8431439995765686, "num_tokens": 218464079.0, "step": 5725 }, { "epoch": 0.7284060552092609, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.684429168701172, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.869590163230896, "num_tokens": 218501307.0, "step": 5726 }, { "epoch": 0.7285332654878515, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.679681777954102, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8688783645629883, "num_tokens": 218537710.0, "step": 5727 }, { "epoch": 0.7286604757664419, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.68623161315918, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8567766547203064, "num_tokens": 218576026.0, "step": 5728 }, { "epoch": 0.7287876860450324, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.701563835144043, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8579884767532349, "num_tokens": 218612491.0, "step": 5729 }, { "epoch": 0.7289148963236229, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.712055206298828, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8697501420974731, "num_tokens": 218642541.0, "step": 5730 }, { "epoch": 0.7290421066022135, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.654632568359375, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8480594158172607, "num_tokens": 218678514.0, "step": 5731 }, { "epoch": 0.729169316880804, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.699037551879883, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8655815720558167, "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": 10.66181755065918, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8689389228820801, "num_tokens": 218754422.0, "step": 5733 }, { "epoch": 0.7294237374379849, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.713155746459961, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.852118968963623, "num_tokens": 218787141.0, "step": 5734 }, { "epoch": 0.7295509477165755, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.670368194580078, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8762990236282349, "num_tokens": 218824764.0, "step": 5735 }, { "epoch": 0.729678157995166, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.75020694732666, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8595396876335144, "num_tokens": 218866142.0, "step": 5736 }, { "epoch": 0.7298053682737565, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.671576499938965, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8563176393508911, "num_tokens": 218909208.0, "step": 5737 }, { "epoch": 0.7299325785523471, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.73254680633545, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8689881563186646, "num_tokens": 218942469.0, "step": 5738 }, { "epoch": 0.7300597888309376, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.696615219116211, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8539155125617981, "num_tokens": 218982934.0, "step": 5739 }, { "epoch": 0.730186999109528, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.707528114318848, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8591592311859131, "num_tokens": 219027055.0, "step": 5740 }, { "epoch": 0.7303142093881185, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.695876121520996, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8614318370819092, "num_tokens": 219069639.0, "step": 5741 }, { "epoch": 0.7304414196667091, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.746395111083984, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8550640344619751, "num_tokens": 219111774.0, "step": 5742 }, { "epoch": 0.7305686299452996, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.668343544006348, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.847719132900238, "num_tokens": 219156962.0, "step": 5743 }, { "epoch": 0.7306958402238901, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.69190502166748, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8697963953018188, "num_tokens": 219193352.0, "step": 5744 }, { "epoch": 0.7308230505024806, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.726283073425293, "learning_rate": 1e-06, "loss": 0.4793, "mean_token_accuracy": 0.846750795841217, "num_tokens": 219230041.0, "step": 5745 }, { "epoch": 0.7309502607810711, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.749861717224121, "learning_rate": 1e-06, "loss": 0.4981, "mean_token_accuracy": 0.8388041257858276, "num_tokens": 219262748.0, "step": 5746 }, { "epoch": 0.7310774710596616, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.741808891296387, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8722052574157715, "num_tokens": 219296556.0, "step": 5747 }, { "epoch": 0.7312046813382521, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.69810962677002, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8477636575698853, "num_tokens": 219330801.0, "step": 5748 }, { "epoch": 0.7313318916168426, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.72810173034668, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8613840341567993, "num_tokens": 219367470.0, "step": 5749 }, { "epoch": 0.7314591018954332, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.706918716430664, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8484402894973755, "num_tokens": 219405291.0, "step": 5750 }, { "epoch": 0.7315863121740237, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.716039657592773, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8623347282409668, "num_tokens": 219443365.0, "step": 5751 }, { "epoch": 0.7317135224526141, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.687163352966309, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8620671033859253, "num_tokens": 219477524.0, "step": 5752 }, { "epoch": 0.7318407327312046, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.73886489868164, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8525494337081909, "num_tokens": 219514576.0, "step": 5753 }, { "epoch": 0.7319679430097952, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.703482627868652, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8612501621246338, "num_tokens": 219558253.0, "step": 5754 }, { "epoch": 0.7320951532883857, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.696593284606934, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8563871383666992, "num_tokens": 219601156.0, "step": 5755 }, { "epoch": 0.7322223635669762, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.734565734863281, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8505809903144836, "num_tokens": 219638241.0, "step": 5756 }, { "epoch": 0.7323495738455668, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.723410606384277, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8614241480827332, "num_tokens": 219678168.0, "step": 5757 }, { "epoch": 0.7324767841241572, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.673964500427246, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8577425479888916, "num_tokens": 219718115.0, "step": 5758 }, { "epoch": 0.7326039944027477, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.743520736694336, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8600708246231079, "num_tokens": 219754221.0, "step": 5759 }, { "epoch": 0.7327312046813382, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.694130897521973, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8668021559715271, "num_tokens": 219786880.0, "step": 5760 }, { "epoch": 0.7328584149599288, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.710074424743652, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8644108772277832, "num_tokens": 219830585.0, "step": 5761 }, { "epoch": 0.7329856252385193, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.764447212219238, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8642911911010742, "num_tokens": 219867249.0, "step": 5762 }, { "epoch": 0.7331128355171098, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.663042068481445, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8525370359420776, "num_tokens": 219906046.0, "step": 5763 }, { "epoch": 0.7332400457957003, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.73717212677002, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.85197913646698, "num_tokens": 219947466.0, "step": 5764 }, { "epoch": 0.7333672560742908, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.718235969543457, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8630468249320984, "num_tokens": 219985698.0, "step": 5765 }, { "epoch": 0.7334944663528813, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.721356391906738, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8615486025810242, "num_tokens": 220026158.0, "step": 5766 }, { "epoch": 0.7336216766314718, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.69546127319336, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.858254075050354, "num_tokens": 220071849.0, "step": 5767 }, { "epoch": 0.7337488869100623, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.670853614807129, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.867774248123169, "num_tokens": 220110255.0, "step": 5768 }, { "epoch": 0.7338760971886529, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.742219924926758, "learning_rate": 1e-06, "loss": 0.4806, "mean_token_accuracy": 0.8462585806846619, "num_tokens": 220144962.0, "step": 5769 }, { "epoch": 0.7340033074672434, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.716987609863281, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8517971634864807, "num_tokens": 220182118.0, "step": 5770 }, { "epoch": 0.7341305177458338, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.69064712524414, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8522766828536987, "num_tokens": 220218954.0, "step": 5771 }, { "epoch": 0.7342577280244243, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.724861145019531, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8591452836990356, "num_tokens": 220259116.0, "step": 5772 }, { "epoch": 0.7343849383030149, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.653877258300781, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8571102023124695, "num_tokens": 220301742.0, "step": 5773 }, { "epoch": 0.7345121485816054, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.689739227294922, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8648989796638489, "num_tokens": 220342572.0, "step": 5774 }, { "epoch": 0.7346393588601959, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.689739227294922, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8662897348403931, "num_tokens": 220376373.0, "step": 5775 }, { "epoch": 0.7347665691387865, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.679680824279785, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8772670030593872, "num_tokens": 220413290.0, "step": 5776 }, { "epoch": 0.7348937794173769, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.72584056854248, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.856282114982605, "num_tokens": 220453219.0, "step": 5777 }, { "epoch": 0.7350209896959674, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.714668273925781, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.861505389213562, "num_tokens": 220488369.0, "step": 5778 }, { "epoch": 0.7351481999745579, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.759352684020996, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8643509745597839, "num_tokens": 220527798.0, "step": 5779 }, { "epoch": 0.7352754102531485, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.719627380371094, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8773884177207947, "num_tokens": 220565398.0, "step": 5780 }, { "epoch": 0.735402620531739, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 10.697430610656738, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8703269958496094, "num_tokens": 220603903.0, "step": 5781 }, { "epoch": 0.7355298308103295, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.71374797821045, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8426073789596558, "num_tokens": 220644168.0, "step": 5782 }, { "epoch": 0.7356570410889199, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.692974090576172, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8728548884391785, "num_tokens": 220679709.0, "step": 5783 }, { "epoch": 0.7357842513675105, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.70361328125, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8548333048820496, "num_tokens": 220721184.0, "step": 5784 }, { "epoch": 0.735911461646101, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.748485565185547, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.865090012550354, "num_tokens": 220755027.0, "step": 5785 }, { "epoch": 0.7360386719246915, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.687278747558594, "learning_rate": 1e-06, "loss": 0.5039, "mean_token_accuracy": 0.8375109434127808, "num_tokens": 220793597.0, "step": 5786 }, { "epoch": 0.736165882203282, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.74744701385498, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8597279787063599, "num_tokens": 220833549.0, "step": 5787 }, { "epoch": 0.7362930924818726, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.675738334655762, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8664074540138245, "num_tokens": 220865551.0, "step": 5788 }, { "epoch": 0.736420302760463, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.69282054901123, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8726069927215576, "num_tokens": 220911626.0, "step": 5789 }, { "epoch": 0.7365475130390535, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.723650932312012, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8615873456001282, "num_tokens": 220949839.0, "step": 5790 }, { "epoch": 0.736674723317644, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.728853225708008, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8442515134811401, "num_tokens": 220984069.0, "step": 5791 }, { "epoch": 0.7368019335962346, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.667119026184082, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8811360001564026, "num_tokens": 221020999.0, "step": 5792 }, { "epoch": 0.7369291438748251, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.752618789672852, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8513448238372803, "num_tokens": 221051082.0, "step": 5793 }, { "epoch": 0.7370563541534156, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.685891151428223, "learning_rate": 1e-06, "loss": 0.4908, "mean_token_accuracy": 0.843941867351532, "num_tokens": 221095444.0, "step": 5794 }, { "epoch": 0.737183564432006, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.717068672180176, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8628718256950378, "num_tokens": 221136433.0, "step": 5795 }, { "epoch": 0.7373107747105966, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.728169441223145, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8528043031692505, "num_tokens": 221171278.0, "step": 5796 }, { "epoch": 0.7374379849891871, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.730854034423828, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8591099381446838, "num_tokens": 221209380.0, "step": 5797 }, { "epoch": 0.7375651952677776, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.677657127380371, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8589319586753845, "num_tokens": 221253076.0, "step": 5798 }, { "epoch": 0.7376924055463682, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.767354965209961, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8557511568069458, "num_tokens": 221290348.0, "step": 5799 }, { "epoch": 0.7378196158249587, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.683717727661133, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8669530153274536, "num_tokens": 221329635.0, "step": 5800 }, { "epoch": 0.7379468261035491, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.783370971679688, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8675509691238403, "num_tokens": 221368521.0, "step": 5801 }, { "epoch": 0.7380740363821396, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.769356727600098, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8733116388320923, "num_tokens": 221407090.0, "step": 5802 }, { "epoch": 0.7382012466607302, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.671587944030762, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.869357168674469, "num_tokens": 221450340.0, "step": 5803 }, { "epoch": 0.7383284569393207, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.722686767578125, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8605896234512329, "num_tokens": 221491996.0, "step": 5804 }, { "epoch": 0.7384556672179112, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.803991317749023, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8694846630096436, "num_tokens": 221531466.0, "step": 5805 }, { "epoch": 0.7385828774965018, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.727550506591797, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.8419580459594727, "num_tokens": 221567003.0, "step": 5806 }, { "epoch": 0.7387100877750922, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.698895454406738, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8736836910247803, "num_tokens": 221607396.0, "step": 5807 }, { "epoch": 0.7388372980536827, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.762809753417969, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.854082465171814, "num_tokens": 221644776.0, "step": 5808 }, { "epoch": 0.7389645083322732, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.726849555969238, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8727520108222961, "num_tokens": 221685643.0, "step": 5809 }, { "epoch": 0.7390917186108638, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.759732246398926, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8541258573532104, "num_tokens": 221724673.0, "step": 5810 }, { "epoch": 0.7392189288894543, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.70942497253418, "learning_rate": 1e-06, "loss": 0.5165, "mean_token_accuracy": 0.8365861773490906, "num_tokens": 221766287.0, "step": 5811 }, { "epoch": 0.7393461391680448, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.724763870239258, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8595684766769409, "num_tokens": 221805357.0, "step": 5812 }, { "epoch": 0.7394733494466353, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.767629623413086, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8581644296646118, "num_tokens": 221838513.0, "step": 5813 }, { "epoch": 0.7396005597252258, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.724390029907227, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8502005338668823, "num_tokens": 221882808.0, "step": 5814 }, { "epoch": 0.7397277700038163, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.68940258026123, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8739176988601685, "num_tokens": 221918505.0, "step": 5815 }, { "epoch": 0.7398549802824068, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.717612266540527, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8621485233306885, "num_tokens": 221955983.0, "step": 5816 }, { "epoch": 0.7399821905609973, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.724661827087402, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8596130013465881, "num_tokens": 221997363.0, "step": 5817 }, { "epoch": 0.7401094008395879, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.759466171264648, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.8462689518928528, "num_tokens": 222039302.0, "step": 5818 }, { "epoch": 0.7402366111181784, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.807823181152344, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8600242137908936, "num_tokens": 222075779.0, "step": 5819 }, { "epoch": 0.7403638213967688, "ewc_loss": 0.020751953125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.68243408203125, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.878612220287323, "num_tokens": 222115120.0, "step": 5820 }, { "epoch": 0.7404910316753593, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.75978946685791, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8526934385299683, "num_tokens": 222156570.0, "step": 5821 }, { "epoch": 0.7406182419539499, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.7545804977417, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.860426664352417, "num_tokens": 222189487.0, "step": 5822 }, { "epoch": 0.7407454522325404, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.760136604309082, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8582587242126465, "num_tokens": 222223163.0, "step": 5823 }, { "epoch": 0.7408726625111309, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.70263957977295, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8738721609115601, "num_tokens": 222259042.0, "step": 5824 }, { "epoch": 0.7409998727897215, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.736394882202148, "learning_rate": 1e-06, "loss": 0.5064, "mean_token_accuracy": 0.8416531085968018, "num_tokens": 222300497.0, "step": 5825 }, { "epoch": 0.7411270830683119, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.777813911437988, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8564540147781372, "num_tokens": 222333630.0, "step": 5826 }, { "epoch": 0.7412542933469024, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.762552261352539, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8560799956321716, "num_tokens": 222371715.0, "step": 5827 }, { "epoch": 0.7413815036254929, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.72377872467041, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8611966967582703, "num_tokens": 222412044.0, "step": 5828 }, { "epoch": 0.7415087139040835, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.753079414367676, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8633488416671753, "num_tokens": 222452668.0, "step": 5829 }, { "epoch": 0.741635924182674, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.756330490112305, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8620858192443848, "num_tokens": 222485631.0, "step": 5830 }, { "epoch": 0.7417631344612645, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.775909423828125, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8756834268569946, "num_tokens": 222522967.0, "step": 5831 }, { "epoch": 0.7418903447398549, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.737868309020996, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8501993417739868, "num_tokens": 222563258.0, "step": 5832 }, { "epoch": 0.7420175550184455, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.756979942321777, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8544266819953918, "num_tokens": 222599248.0, "step": 5833 }, { "epoch": 0.742144765297036, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.757247924804688, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8677937388420105, "num_tokens": 222642575.0, "step": 5834 }, { "epoch": 0.7422719755756265, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.745878219604492, "learning_rate": 1e-06, "loss": 0.5015, "mean_token_accuracy": 0.837040364742279, "num_tokens": 222687330.0, "step": 5835 }, { "epoch": 0.742399185854217, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.826436996459961, "learning_rate": 1e-06, "loss": 0.5172, "mean_token_accuracy": 0.8355711102485657, "num_tokens": 222720020.0, "step": 5836 }, { "epoch": 0.7425263961328076, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.758627891540527, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8744411468505859, "num_tokens": 222758634.0, "step": 5837 }, { "epoch": 0.742653606411398, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.760539054870605, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8677716255187988, "num_tokens": 222794823.0, "step": 5838 }, { "epoch": 0.7427808166899885, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.712809562683105, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.860265851020813, "num_tokens": 222833836.0, "step": 5839 }, { "epoch": 0.742908026968579, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.726533889770508, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8619062900543213, "num_tokens": 222875712.0, "step": 5840 }, { "epoch": 0.7430352372471696, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.731287002563477, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8498016595840454, "num_tokens": 222915494.0, "step": 5841 }, { "epoch": 0.7431624475257601, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.724453926086426, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8768866062164307, "num_tokens": 222951675.0, "step": 5842 }, { "epoch": 0.7432896578043506, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.747065544128418, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8550005555152893, "num_tokens": 222993245.0, "step": 5843 }, { "epoch": 0.743416868082941, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.745457649230957, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8669159412384033, "num_tokens": 223029876.0, "step": 5844 }, { "epoch": 0.7435440783615316, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.714739799499512, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8661800622940063, "num_tokens": 223072550.0, "step": 5845 }, { "epoch": 0.7436712886401221, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.754922866821289, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8594310879707336, "num_tokens": 223112892.0, "step": 5846 }, { "epoch": 0.7437984989187126, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.71152400970459, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.865973711013794, "num_tokens": 223157093.0, "step": 5847 }, { "epoch": 0.7439257091973032, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.80765438079834, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8516314029693604, "num_tokens": 223190756.0, "step": 5848 }, { "epoch": 0.7440529194758937, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.798080444335938, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8523569107055664, "num_tokens": 223226013.0, "step": 5849 }, { "epoch": 0.7441801297544841, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.808716773986816, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8733636140823364, "num_tokens": 223262858.0, "step": 5850 }, { "epoch": 0.7443073400330746, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.708972930908203, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8643984794616699, "num_tokens": 223310528.0, "step": 5851 }, { "epoch": 0.7444345503116652, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.792679786682129, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8633522987365723, "num_tokens": 223345021.0, "step": 5852 }, { "epoch": 0.7445617605902557, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.764921188354492, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8710812330245972, "num_tokens": 223382055.0, "step": 5853 }, { "epoch": 0.7446889708688462, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.786970138549805, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8502514362335205, "num_tokens": 223418441.0, "step": 5854 }, { "epoch": 0.7448161811474368, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.765691757202148, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.858814001083374, "num_tokens": 223458150.0, "step": 5855 }, { "epoch": 0.7449433914260272, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.746524810791016, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8565937280654907, "num_tokens": 223496916.0, "step": 5856 }, { "epoch": 0.7450706017046177, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.774175643920898, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8662223815917969, "num_tokens": 223534808.0, "step": 5857 }, { "epoch": 0.7451978119832082, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.773058891296387, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8694994449615479, "num_tokens": 223571255.0, "step": 5858 }, { "epoch": 0.7453250222617988, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.750385284423828, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8641345500946045, "num_tokens": 223616608.0, "step": 5859 }, { "epoch": 0.7454522325403893, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.7427396774292, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8604123592376709, "num_tokens": 223656188.0, "step": 5860 }, { "epoch": 0.7455794428189798, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.749984741210938, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.855979323387146, "num_tokens": 223693787.0, "step": 5861 }, { "epoch": 0.7457066530975703, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.773746490478516, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8717815279960632, "num_tokens": 223729350.0, "step": 5862 }, { "epoch": 0.7458338633761608, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.75172233581543, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.848567545413971, "num_tokens": 223770846.0, "step": 5863 }, { "epoch": 0.7459610736547513, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.742034912109375, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8600081205368042, "num_tokens": 223814194.0, "step": 5864 }, { "epoch": 0.7460882839333418, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.809209823608398, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8537185788154602, "num_tokens": 223849141.0, "step": 5865 }, { "epoch": 0.7462154942119323, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.727697372436523, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8819437026977539, "num_tokens": 223888654.0, "step": 5866 }, { "epoch": 0.7463427044905229, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.7439546585083, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8652843832969666, "num_tokens": 223924509.0, "step": 5867 }, { "epoch": 0.7464699147691134, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.708605766296387, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8463387489318848, "num_tokens": 223966191.0, "step": 5868 }, { "epoch": 0.7465971250477038, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.759430885314941, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8628677725791931, "num_tokens": 224001415.0, "step": 5869 }, { "epoch": 0.7467243353262943, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.732024192810059, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8662605285644531, "num_tokens": 224042820.0, "step": 5870 }, { "epoch": 0.7468515456048849, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.767669677734375, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8660412430763245, "num_tokens": 224078411.0, "step": 5871 }, { "epoch": 0.7469787558834754, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.726292610168457, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8674448132514954, "num_tokens": 224117136.0, "step": 5872 }, { "epoch": 0.7471059661620659, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 10.762273788452148, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8550136089324951, "num_tokens": 224155879.0, "step": 5873 }, { "epoch": 0.7472331764406565, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.744100570678711, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8728472590446472, "num_tokens": 224195186.0, "step": 5874 }, { "epoch": 0.7473603867192469, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.752464294433594, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.858756422996521, "num_tokens": 224235722.0, "step": 5875 }, { "epoch": 0.7474875969978374, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.752679824829102, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8683103919029236, "num_tokens": 224268461.0, "step": 5876 }, { "epoch": 0.7476148072764279, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.736690521240234, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8737661242485046, "num_tokens": 224302921.0, "step": 5877 }, { "epoch": 0.7477420175550185, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.712686538696289, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8620063066482544, "num_tokens": 224344081.0, "step": 5878 }, { "epoch": 0.747869227833609, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.743073463439941, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8494914770126343, "num_tokens": 224384464.0, "step": 5879 }, { "epoch": 0.7479964381121995, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.679112434387207, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.869382381439209, "num_tokens": 224425542.0, "step": 5880 }, { "epoch": 0.7481236483907899, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.741294860839844, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8780428171157837, "num_tokens": 224465342.0, "step": 5881 }, { "epoch": 0.7482508586693805, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.75909423828125, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8662831783294678, "num_tokens": 224504787.0, "step": 5882 }, { "epoch": 0.748378068947971, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.70500373840332, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8619798421859741, "num_tokens": 224548534.0, "step": 5883 }, { "epoch": 0.7485052792265615, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.723121643066406, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8535892963409424, "num_tokens": 224590251.0, "step": 5884 }, { "epoch": 0.748632489505152, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.729716300964355, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8715811371803284, "num_tokens": 224633745.0, "step": 5885 }, { "epoch": 0.7487596997837426, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 10.724661827087402, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.846564769744873, "num_tokens": 224671319.0, "step": 5886 }, { "epoch": 0.748886910062333, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.780781745910645, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8612127900123596, "num_tokens": 224712670.0, "step": 5887 }, { "epoch": 0.7490141203409235, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.699003219604492, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8596916794776917, "num_tokens": 224747840.0, "step": 5888 }, { "epoch": 0.749141330619514, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.7468900680542, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.870993435382843, "num_tokens": 224783274.0, "step": 5889 }, { "epoch": 0.7492685408981046, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.751885414123535, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8564141392707825, "num_tokens": 224820456.0, "step": 5890 }, { "epoch": 0.7493957511766951, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.692733764648438, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8659853339195251, "num_tokens": 224858569.0, "step": 5891 }, { "epoch": 0.7495229614552856, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.700272560119629, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8597780466079712, "num_tokens": 224901033.0, "step": 5892 }, { "epoch": 0.749650171733876, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.742098808288574, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8756279945373535, "num_tokens": 224934345.0, "step": 5893 }, { "epoch": 0.7497773820124666, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.721879005432129, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8542079925537109, "num_tokens": 224972479.0, "step": 5894 }, { "epoch": 0.7499045922910571, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.718244552612305, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.877685546875, "num_tokens": 225014213.0, "step": 5895 }, { "epoch": 0.7500318025696476, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.73261833190918, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8680540323257446, "num_tokens": 225049929.0, "step": 5896 }, { "epoch": 0.7501590128482382, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.735456466674805, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8454829454421997, "num_tokens": 225086409.0, "step": 5897 }, { "epoch": 0.7502862231268287, "ewc_loss": 0.02099609375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 10.736466407775879, "learning_rate": 1e-06, "loss": 0.4901, "mean_token_accuracy": 0.8435937166213989, "num_tokens": 225124443.0, "step": 5898 }, { "epoch": 0.7504134334054191, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.775739669799805, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8507376909255981, "num_tokens": 225166665.0, "step": 5899 }, { "epoch": 0.7505406436840096, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.74250316619873, "learning_rate": 1e-06, "loss": 0.5125, "mean_token_accuracy": 0.8376061320304871, "num_tokens": 225203841.0, "step": 5900 }, { "epoch": 0.7506678539626002, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.796979904174805, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.8447288274765015, "num_tokens": 225234844.0, "step": 5901 }, { "epoch": 0.7507950642411907, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.831342697143555, "learning_rate": 1e-06, "loss": 0.4899, "mean_token_accuracy": 0.8432350158691406, "num_tokens": 225269828.0, "step": 5902 }, { "epoch": 0.7509222745197812, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.712190628051758, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8654971122741699, "num_tokens": 225312361.0, "step": 5903 }, { "epoch": 0.7510494847983717, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.765763282775879, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8500653505325317, "num_tokens": 225353708.0, "step": 5904 }, { "epoch": 0.7511766950769622, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.736634254455566, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8798163533210754, "num_tokens": 225399730.0, "step": 5905 }, { "epoch": 0.7513039053555527, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.688370704650879, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8548393845558167, "num_tokens": 225441836.0, "step": 5906 }, { "epoch": 0.7514311156341432, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.849955558776855, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8679940700531006, "num_tokens": 225471004.0, "step": 5907 }, { "epoch": 0.7515583259127337, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.783644676208496, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8426311612129211, "num_tokens": 225505789.0, "step": 5908 }, { "epoch": 0.7516855361913243, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.734650611877441, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8629181385040283, "num_tokens": 225541138.0, "step": 5909 }, { "epoch": 0.7518127464699148, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.768424034118652, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8642162680625916, "num_tokens": 225581695.0, "step": 5910 }, { "epoch": 0.7519399567485053, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.732726097106934, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.86726975440979, "num_tokens": 225619532.0, "step": 5911 }, { "epoch": 0.7520671670270958, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.781095504760742, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.852152943611145, "num_tokens": 225657403.0, "step": 5912 }, { "epoch": 0.7521943773056863, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.78682804107666, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8567063808441162, "num_tokens": 225693402.0, "step": 5913 }, { "epoch": 0.7523215875842768, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.752666473388672, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8596706390380859, "num_tokens": 225727208.0, "step": 5914 }, { "epoch": 0.7524487978628673, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.715007781982422, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8655041456222534, "num_tokens": 225762117.0, "step": 5915 }, { "epoch": 0.7525760081414579, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.743171691894531, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8459752798080444, "num_tokens": 225794737.0, "step": 5916 }, { "epoch": 0.7527032184200484, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.7343111038208, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8598717451095581, "num_tokens": 225829074.0, "step": 5917 }, { "epoch": 0.7528304286986388, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.76488208770752, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8735485076904297, "num_tokens": 225871344.0, "step": 5918 }, { "epoch": 0.7529576389772293, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.772967338562012, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8557724952697754, "num_tokens": 225911321.0, "step": 5919 }, { "epoch": 0.7530848492558199, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.735673904418945, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8687955141067505, "num_tokens": 225944613.0, "step": 5920 }, { "epoch": 0.7532120595344104, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.753021240234375, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8654392957687378, "num_tokens": 225989501.0, "step": 5921 }, { "epoch": 0.7533392698130009, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.765308380126953, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8625134229660034, "num_tokens": 226024984.0, "step": 5922 }, { "epoch": 0.7534664800915915, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.762085914611816, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.874250054359436, "num_tokens": 226063255.0, "step": 5923 }, { "epoch": 0.7535936903701819, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.752307891845703, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8486470580101013, "num_tokens": 226105538.0, "step": 5924 }, { "epoch": 0.7537209006487724, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.80025577545166, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8645567893981934, "num_tokens": 226137449.0, "step": 5925 }, { "epoch": 0.7538481109273629, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.744874954223633, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8544340133666992, "num_tokens": 226171078.0, "step": 5926 }, { "epoch": 0.7539753212059535, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.781830787658691, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8621258735656738, "num_tokens": 226208060.0, "step": 5927 }, { "epoch": 0.754102531484544, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.694341659545898, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8713436126708984, "num_tokens": 226245330.0, "step": 5928 }, { "epoch": 0.7542297417631345, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.794910430908203, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8819692134857178, "num_tokens": 226281974.0, "step": 5929 }, { "epoch": 0.7543569520417249, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.703743934631348, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8512179851531982, "num_tokens": 226323302.0, "step": 5930 }, { "epoch": 0.7544841623203155, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.741777420043945, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8483706116676331, "num_tokens": 226361346.0, "step": 5931 }, { "epoch": 0.754611372598906, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.777302742004395, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8567895293235779, "num_tokens": 226397744.0, "step": 5932 }, { "epoch": 0.7547385828774965, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.785602569580078, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8686479330062866, "num_tokens": 226436745.0, "step": 5933 }, { "epoch": 0.754865793156087, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.718015670776367, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.86948561668396, "num_tokens": 226470302.0, "step": 5934 }, { "epoch": 0.7549930034346776, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.870742797851562, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.863632082939148, "num_tokens": 226500090.0, "step": 5935 }, { "epoch": 0.755120213713268, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.81038761138916, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8628732562065125, "num_tokens": 226533359.0, "step": 5936 }, { "epoch": 0.7552474239918585, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.78327751159668, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8519028425216675, "num_tokens": 226567060.0, "step": 5937 }, { "epoch": 0.755374634270449, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.83322525024414, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8592984676361084, "num_tokens": 226604270.0, "step": 5938 }, { "epoch": 0.7555018445490396, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.757579803466797, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8595386743545532, "num_tokens": 226641310.0, "step": 5939 }, { "epoch": 0.7556290548276301, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.756317138671875, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8503301739692688, "num_tokens": 226678305.0, "step": 5940 }, { "epoch": 0.7557562651062206, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.77575397491455, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8677314519882202, "num_tokens": 226715628.0, "step": 5941 }, { "epoch": 0.755883475384811, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.767549514770508, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8730870485305786, "num_tokens": 226755346.0, "step": 5942 }, { "epoch": 0.7560106856634016, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.826611518859863, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8503807783126831, "num_tokens": 226792243.0, "step": 5943 }, { "epoch": 0.7561378959419921, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.810837745666504, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.874557375907898, "num_tokens": 226829033.0, "step": 5944 }, { "epoch": 0.7562651062205826, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.754964828491211, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8761414885520935, "num_tokens": 226870645.0, "step": 5945 }, { "epoch": 0.7563923164991732, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.796985626220703, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8656847476959229, "num_tokens": 226902395.0, "step": 5946 }, { "epoch": 0.7565195267777637, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.728187561035156, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8677024841308594, "num_tokens": 226942620.0, "step": 5947 }, { "epoch": 0.7566467370563541, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.77103328704834, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8687381744384766, "num_tokens": 226976472.0, "step": 5948 }, { "epoch": 0.7567739473349446, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.833183288574219, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8521556258201599, "num_tokens": 227013626.0, "step": 5949 }, { "epoch": 0.7569011576135352, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.870620727539062, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8484214544296265, "num_tokens": 227049336.0, "step": 5950 }, { "epoch": 0.7570283678921257, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.755240440368652, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8842753767967224, "num_tokens": 227087371.0, "step": 5951 }, { "epoch": 0.7571555781707162, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.738734245300293, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8635044097900391, "num_tokens": 227129188.0, "step": 5952 }, { "epoch": 0.7572827884493067, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.773313522338867, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8544468283653259, "num_tokens": 227161807.0, "step": 5953 }, { "epoch": 0.7574099987278972, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.766629219055176, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8530398607254028, "num_tokens": 227198431.0, "step": 5954 }, { "epoch": 0.7575372090064877, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.716175079345703, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8816220164299011, "num_tokens": 227236046.0, "step": 5955 }, { "epoch": 0.7576644192850782, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.742712020874023, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8774515390396118, "num_tokens": 227272925.0, "step": 5956 }, { "epoch": 0.7577916295636687, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.819931983947754, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8657777309417725, "num_tokens": 227312311.0, "step": 5957 }, { "epoch": 0.7579188398422593, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.745174407958984, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8820255994796753, "num_tokens": 227349632.0, "step": 5958 }, { "epoch": 0.7580460501208498, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.763312339782715, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.859784722328186, "num_tokens": 227392982.0, "step": 5959 }, { "epoch": 0.7581732603994403, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.793537139892578, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8588176965713501, "num_tokens": 227431878.0, "step": 5960 }, { "epoch": 0.7583004706780307, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.752044677734375, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8515439033508301, "num_tokens": 227470401.0, "step": 5961 }, { "epoch": 0.7584276809566213, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.783042907714844, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8668872117996216, "num_tokens": 227507189.0, "step": 5962 }, { "epoch": 0.7585548912352118, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.737883567810059, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.865837812423706, "num_tokens": 227548070.0, "step": 5963 }, { "epoch": 0.7586821015138023, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.722118377685547, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.867477297782898, "num_tokens": 227584360.0, "step": 5964 }, { "epoch": 0.7588093117923929, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.783774375915527, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8687855005264282, "num_tokens": 227618961.0, "step": 5965 }, { "epoch": 0.7589365220709834, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.729854583740234, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8629069328308105, "num_tokens": 227658304.0, "step": 5966 }, { "epoch": 0.7590637323495738, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.77915096282959, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8604371547698975, "num_tokens": 227701526.0, "step": 5967 }, { "epoch": 0.7591909426281643, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.731623649597168, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8678325414657593, "num_tokens": 227744930.0, "step": 5968 }, { "epoch": 0.7593181529067549, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.729639053344727, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8729845881462097, "num_tokens": 227780607.0, "step": 5969 }, { "epoch": 0.7594453631853454, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.779138565063477, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.858749270439148, "num_tokens": 227825742.0, "step": 5970 }, { "epoch": 0.7595725734639359, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.693022727966309, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8433187007904053, "num_tokens": 227865278.0, "step": 5971 }, { "epoch": 0.7596997837425264, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.768523216247559, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.858817458152771, "num_tokens": 227906912.0, "step": 5972 }, { "epoch": 0.7598269940211169, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.781111717224121, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8601374626159668, "num_tokens": 227947067.0, "step": 5973 }, { "epoch": 0.7599542042997074, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.744351387023926, "learning_rate": 1e-06, "loss": 0.4867, "mean_token_accuracy": 0.8443047404289246, "num_tokens": 227988563.0, "step": 5974 }, { "epoch": 0.7600814145782979, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.748635292053223, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8546386361122131, "num_tokens": 228028556.0, "step": 5975 }, { "epoch": 0.7602086248568884, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.735963821411133, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8770655989646912, "num_tokens": 228066265.0, "step": 5976 }, { "epoch": 0.760335835135479, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.808432579040527, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.858791172504425, "num_tokens": 228103732.0, "step": 5977 }, { "epoch": 0.7604630454140695, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.725287437438965, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8601545095443726, "num_tokens": 228149609.0, "step": 5978 }, { "epoch": 0.7605902556926599, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.735962867736816, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8645716905593872, "num_tokens": 228188462.0, "step": 5979 }, { "epoch": 0.7607174659712505, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 10.74333381652832, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.875078558921814, "num_tokens": 228229252.0, "step": 5980 }, { "epoch": 0.760844676249841, "ewc_loss": 0.021240234375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.771023750305176, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8599042296409607, "num_tokens": 228267397.0, "step": 5981 }, { "epoch": 0.7609718865284315, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.766831398010254, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8703150153160095, "num_tokens": 228307561.0, "step": 5982 }, { "epoch": 0.761099096807022, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.799545288085938, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8643198609352112, "num_tokens": 228347667.0, "step": 5983 }, { "epoch": 0.7612263070856126, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.772486686706543, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8596919775009155, "num_tokens": 228386820.0, "step": 5984 }, { "epoch": 0.761353517364203, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.808923721313477, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8716432452201843, "num_tokens": 228427304.0, "step": 5985 }, { "epoch": 0.7614807276427935, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.762639999389648, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8527992963790894, "num_tokens": 228470779.0, "step": 5986 }, { "epoch": 0.761607937921384, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.766694068908691, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.878334641456604, "num_tokens": 228500454.0, "step": 5987 }, { "epoch": 0.7617351481999746, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.816085815429688, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8704161047935486, "num_tokens": 228537374.0, "step": 5988 }, { "epoch": 0.7618623584785651, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.771956443786621, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8630602359771729, "num_tokens": 228577215.0, "step": 5989 }, { "epoch": 0.7619895687571556, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.757553100585938, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8672165870666504, "num_tokens": 228611693.0, "step": 5990 }, { "epoch": 0.762116779035746, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.753458976745605, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8611834645271301, "num_tokens": 228644231.0, "step": 5991 }, { "epoch": 0.7622439893143366, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.707571029663086, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8734678626060486, "num_tokens": 228684776.0, "step": 5992 }, { "epoch": 0.7623711995929271, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.70087718963623, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8682260513305664, "num_tokens": 228724214.0, "step": 5993 }, { "epoch": 0.7624984098715176, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.792074203491211, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8602380752563477, "num_tokens": 228765098.0, "step": 5994 }, { "epoch": 0.7626256201501082, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.748908042907715, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8638632893562317, "num_tokens": 228802423.0, "step": 5995 }, { "epoch": 0.7627528304286987, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.779922485351562, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8516457080841064, "num_tokens": 228840218.0, "step": 5996 }, { "epoch": 0.7628800407072891, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.7388277053833, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8673256635665894, "num_tokens": 228882810.0, "step": 5997 }, { "epoch": 0.7630072509858796, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.790351867675781, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8697454333305359, "num_tokens": 228917664.0, "step": 5998 }, { "epoch": 0.7631344612644702, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.752767562866211, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8687554597854614, "num_tokens": 228951017.0, "step": 5999 }, { "epoch": 0.7632616715430607, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.77007007598877, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8643378019332886, "num_tokens": 228990949.0, "step": 6000 }, { "epoch": 0.7633888818216512, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.794465065002441, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8681495189666748, "num_tokens": 229026604.0, "step": 6001 }, { "epoch": 0.7635160921002417, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.762372016906738, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.856526792049408, "num_tokens": 229061810.0, "step": 6002 }, { "epoch": 0.7636433023788322, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.741155624389648, "learning_rate": 1e-06, "loss": 0.4956, "mean_token_accuracy": 0.8400651216506958, "num_tokens": 229107765.0, "step": 6003 }, { "epoch": 0.7637705126574227, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.759757995605469, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8504984378814697, "num_tokens": 229147822.0, "step": 6004 }, { "epoch": 0.7638977229360132, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.762823104858398, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8724798560142517, "num_tokens": 229184421.0, "step": 6005 }, { "epoch": 0.7640249332146037, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.79970645904541, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8632907867431641, "num_tokens": 229221737.0, "step": 6006 }, { "epoch": 0.7641521434931943, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.820181846618652, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8562494516372681, "num_tokens": 229254727.0, "step": 6007 }, { "epoch": 0.7642793537717848, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.79754638671875, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8734948635101318, "num_tokens": 229292648.0, "step": 6008 }, { "epoch": 0.7644065640503753, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.791925430297852, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8676791787147522, "num_tokens": 229335708.0, "step": 6009 }, { "epoch": 0.7645337743289657, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.814249992370605, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8714459538459778, "num_tokens": 229373791.0, "step": 6010 }, { "epoch": 0.7646609846075563, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.83275318145752, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8731802701950073, "num_tokens": 229409891.0, "step": 6011 }, { "epoch": 0.7647881948861468, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.801178932189941, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8643664121627808, "num_tokens": 229449859.0, "step": 6012 }, { "epoch": 0.7649154051647373, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.811869621276855, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8702056407928467, "num_tokens": 229484224.0, "step": 6013 }, { "epoch": 0.7650426154433279, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.802220344543457, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8583824634552002, "num_tokens": 229521104.0, "step": 6014 }, { "epoch": 0.7651698257219184, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.798077583312988, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8474825620651245, "num_tokens": 229558918.0, "step": 6015 }, { "epoch": 0.7652970360005088, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.806876182556152, "learning_rate": 1e-06, "loss": 0.4811, "mean_token_accuracy": 0.8496600985527039, "num_tokens": 229596931.0, "step": 6016 }, { "epoch": 0.7654242462790993, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.846256256103516, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8479900360107422, "num_tokens": 229633903.0, "step": 6017 }, { "epoch": 0.7655514565576899, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.774252891540527, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.859808087348938, "num_tokens": 229674224.0, "step": 6018 }, { "epoch": 0.7656786668362804, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.802298545837402, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8615007400512695, "num_tokens": 229710547.0, "step": 6019 }, { "epoch": 0.7658058771148709, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.807740211486816, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8779295682907104, "num_tokens": 229752020.0, "step": 6020 }, { "epoch": 0.7659330873934614, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.824503898620605, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8595661520957947, "num_tokens": 229785710.0, "step": 6021 }, { "epoch": 0.7660602976720519, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84060287475586, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8697376847267151, "num_tokens": 229824118.0, "step": 6022 }, { "epoch": 0.7661875079506424, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.814695358276367, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8749549388885498, "num_tokens": 229854231.0, "step": 6023 }, { "epoch": 0.7663147182292329, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.861948013305664, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8530595898628235, "num_tokens": 229888102.0, "step": 6024 }, { "epoch": 0.7664419285078234, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.78189468383789, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8676680326461792, "num_tokens": 229927634.0, "step": 6025 }, { "epoch": 0.766569138786414, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.769926071166992, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8711668848991394, "num_tokens": 229966526.0, "step": 6026 }, { "epoch": 0.7666963490650045, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.798882484436035, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.872705340385437, "num_tokens": 230002761.0, "step": 6027 }, { "epoch": 0.7668235593435949, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.851350784301758, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8679436445236206, "num_tokens": 230041303.0, "step": 6028 }, { "epoch": 0.7669507696221854, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.778328895568848, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8672263622283936, "num_tokens": 230081117.0, "step": 6029 }, { "epoch": 0.767077979900776, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.758694648742676, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8585389852523804, "num_tokens": 230119411.0, "step": 6030 }, { "epoch": 0.7672051901793665, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.818925857543945, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8703572750091553, "num_tokens": 230161111.0, "step": 6031 }, { "epoch": 0.767332400457957, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.795820236206055, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8585917353630066, "num_tokens": 230196999.0, "step": 6032 }, { "epoch": 0.7674596107365476, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.823312759399414, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8653165698051453, "num_tokens": 230231495.0, "step": 6033 }, { "epoch": 0.767586821015138, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.779060363769531, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8566493391990662, "num_tokens": 230267732.0, "step": 6034 }, { "epoch": 0.7677140312937285, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.851617813110352, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8623692989349365, "num_tokens": 230297559.0, "step": 6035 }, { "epoch": 0.767841241572319, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.761804580688477, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.851421058177948, "num_tokens": 230336396.0, "step": 6036 }, { "epoch": 0.7679684518509096, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.81246280670166, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8620851039886475, "num_tokens": 230368152.0, "step": 6037 }, { "epoch": 0.7680956621295001, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.792060852050781, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8678675889968872, "num_tokens": 230408081.0, "step": 6038 }, { "epoch": 0.7682228724080906, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.813294410705566, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.855713963508606, "num_tokens": 230441691.0, "step": 6039 }, { "epoch": 0.768350082686681, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.774945259094238, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8612985610961914, "num_tokens": 230480089.0, "step": 6040 }, { "epoch": 0.7684772929652716, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.802040100097656, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8574180603027344, "num_tokens": 230518655.0, "step": 6041 }, { "epoch": 0.7686045032438621, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84170150756836, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8660828471183777, "num_tokens": 230555717.0, "step": 6042 }, { "epoch": 0.7687317135224526, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.807236671447754, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8712384700775146, "num_tokens": 230596699.0, "step": 6043 }, { "epoch": 0.7688589238010431, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.773873329162598, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8718593120574951, "num_tokens": 230633284.0, "step": 6044 }, { "epoch": 0.7689861340796337, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.830147743225098, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8662713170051575, "num_tokens": 230673133.0, "step": 6045 }, { "epoch": 0.7691133443582241, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.821236610412598, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8639989495277405, "num_tokens": 230719003.0, "step": 6046 }, { "epoch": 0.7692405546368146, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.747146606445312, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8720404505729675, "num_tokens": 230754805.0, "step": 6047 }, { "epoch": 0.7693677649154052, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.868762969970703, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8501101136207581, "num_tokens": 230793249.0, "step": 6048 }, { "epoch": 0.7694949751939957, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.80599308013916, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8806031346321106, "num_tokens": 230826025.0, "step": 6049 }, { "epoch": 0.7696221854725862, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.749735832214355, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8748705387115479, "num_tokens": 230866615.0, "step": 6050 }, { "epoch": 0.7697493957511767, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.882425308227539, "learning_rate": 1e-06, "loss": 0.4791, "mean_token_accuracy": 0.8483837842941284, "num_tokens": 230901587.0, "step": 6051 }, { "epoch": 0.7698766060297672, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.789196014404297, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8684921860694885, "num_tokens": 230944267.0, "step": 6052 }, { "epoch": 0.7700038163083577, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.834451675415039, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8540099859237671, "num_tokens": 230980022.0, "step": 6053 }, { "epoch": 0.7701310265869482, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.809977531433105, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8692700862884521, "num_tokens": 231018110.0, "step": 6054 }, { "epoch": 0.7702582368655387, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.817424774169922, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8685289025306702, "num_tokens": 231056531.0, "step": 6055 }, { "epoch": 0.7703854471441293, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.8002347946167, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.869110107421875, "num_tokens": 231096270.0, "step": 6056 }, { "epoch": 0.7705126574227198, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.79472827911377, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8563389778137207, "num_tokens": 231136300.0, "step": 6057 }, { "epoch": 0.7706398677013102, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.804584503173828, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.849892795085907, "num_tokens": 231169005.0, "step": 6058 }, { "epoch": 0.7707670779799007, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.794696807861328, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8699690103530884, "num_tokens": 231205274.0, "step": 6059 }, { "epoch": 0.7708942882584913, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.803739547729492, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8683258295059204, "num_tokens": 231239022.0, "step": 6060 }, { "epoch": 0.7710214985370818, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.816452026367188, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.844971776008606, "num_tokens": 231282952.0, "step": 6061 }, { "epoch": 0.7711487088156723, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.785832405090332, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8613705635070801, "num_tokens": 231322980.0, "step": 6062 }, { "epoch": 0.7712759190942629, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.8333158493042, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8549253940582275, "num_tokens": 231358382.0, "step": 6063 }, { "epoch": 0.7714031293728534, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.764972686767578, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8708745837211609, "num_tokens": 231401915.0, "step": 6064 }, { "epoch": 0.7715303396514438, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.853205680847168, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8411862254142761, "num_tokens": 231433241.0, "step": 6065 }, { "epoch": 0.7716575499300343, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.803414344787598, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8574502468109131, "num_tokens": 231478318.0, "step": 6066 }, { "epoch": 0.7717847602086249, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.794865608215332, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8555845022201538, "num_tokens": 231518584.0, "step": 6067 }, { "epoch": 0.7719119704872154, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.824455261230469, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.862714409828186, "num_tokens": 231556689.0, "step": 6068 }, { "epoch": 0.7720391807658059, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.794965744018555, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8621149063110352, "num_tokens": 231593936.0, "step": 6069 }, { "epoch": 0.7721663910443964, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.861288070678711, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8821049928665161, "num_tokens": 231634785.0, "step": 6070 }, { "epoch": 0.7722936013229869, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.805179595947266, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8587364554405212, "num_tokens": 231675798.0, "step": 6071 }, { "epoch": 0.7724208116015774, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.818891525268555, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8604539632797241, "num_tokens": 231711611.0, "step": 6072 }, { "epoch": 0.7725480218801679, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.869624137878418, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8529294729232788, "num_tokens": 231746997.0, "step": 6073 }, { "epoch": 0.7726752321587584, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.825044631958008, "learning_rate": 1e-06, "loss": 0.4854, "mean_token_accuracy": 0.8422775864601135, "num_tokens": 231787696.0, "step": 6074 }, { "epoch": 0.772802442437349, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.845320701599121, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8541274666786194, "num_tokens": 231826127.0, "step": 6075 }, { "epoch": 0.7729296527159395, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.837461471557617, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8575973510742188, "num_tokens": 231857538.0, "step": 6076 }, { "epoch": 0.7730568629945299, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.835165977478027, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8604652881622314, "num_tokens": 231893913.0, "step": 6077 }, { "epoch": 0.7731840732731204, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.781621932983398, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.864657461643219, "num_tokens": 231934274.0, "step": 6078 }, { "epoch": 0.773311283551711, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.786772727966309, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8568887710571289, "num_tokens": 231983197.0, "step": 6079 }, { "epoch": 0.7734384938303015, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.798526763916016, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8581018447875977, "num_tokens": 232021265.0, "step": 6080 }, { "epoch": 0.773565704108892, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.820455551147461, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8609827756881714, "num_tokens": 232060992.0, "step": 6081 }, { "epoch": 0.7736929143874826, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.761960983276367, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8600931167602539, "num_tokens": 232096758.0, "step": 6082 }, { "epoch": 0.773820124666073, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84388256072998, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8583788275718689, "num_tokens": 232133773.0, "step": 6083 }, { "epoch": 0.7739473349446635, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.797218322753906, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8602094650268555, "num_tokens": 232173182.0, "step": 6084 }, { "epoch": 0.774074545223254, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.808158874511719, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8704649209976196, "num_tokens": 232206478.0, "step": 6085 }, { "epoch": 0.7742017555018446, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.844901084899902, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8562790155410767, "num_tokens": 232242231.0, "step": 6086 }, { "epoch": 0.7743289657804351, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.80727767944336, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8660496473312378, "num_tokens": 232270636.0, "step": 6087 }, { "epoch": 0.7744561760590256, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.820791244506836, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8677905201911926, "num_tokens": 232309974.0, "step": 6088 }, { "epoch": 0.774583386337616, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.809958457946777, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8719335198402405, "num_tokens": 232344118.0, "step": 6089 }, { "epoch": 0.7747105966162066, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.813116073608398, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8519987463951111, "num_tokens": 232385183.0, "step": 6090 }, { "epoch": 0.7748378068947971, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.779205322265625, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8639557957649231, "num_tokens": 232421670.0, "step": 6091 }, { "epoch": 0.7749650171733876, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.811494827270508, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.850476861000061, "num_tokens": 232459563.0, "step": 6092 }, { "epoch": 0.7750922274519781, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.818528175354004, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8626925349235535, "num_tokens": 232496357.0, "step": 6093 }, { "epoch": 0.7752194377305687, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.758915901184082, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8626642823219299, "num_tokens": 232530436.0, "step": 6094 }, { "epoch": 0.7753466480091591, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.888086318969727, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.856634259223938, "num_tokens": 232568585.0, "step": 6095 }, { "epoch": 0.7754738582877496, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.815914154052734, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8687900304794312, "num_tokens": 232602506.0, "step": 6096 }, { "epoch": 0.7756010685663401, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.723111152648926, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8704703450202942, "num_tokens": 232642819.0, "step": 6097 }, { "epoch": 0.7757282788449307, "ewc_loss": 0.021484375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 10.766641616821289, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8493547439575195, "num_tokens": 232679505.0, "step": 6098 }, { "epoch": 0.7758554891235212, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.800869941711426, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8667091131210327, "num_tokens": 232715701.0, "step": 6099 }, { "epoch": 0.7759826994021117, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84029769897461, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8634620904922485, "num_tokens": 232750786.0, "step": 6100 }, { "epoch": 0.7761099096807021, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.796780586242676, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8507590293884277, "num_tokens": 232791370.0, "step": 6101 }, { "epoch": 0.7762371199592927, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.8074312210083, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8632298111915588, "num_tokens": 232830895.0, "step": 6102 }, { "epoch": 0.7763643302378832, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.781500816345215, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8611989617347717, "num_tokens": 232869940.0, "step": 6103 }, { "epoch": 0.7764915405164737, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.781193733215332, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8581130504608154, "num_tokens": 232910743.0, "step": 6104 }, { "epoch": 0.7766187507950643, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.809526443481445, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8614357709884644, "num_tokens": 232948428.0, "step": 6105 }, { "epoch": 0.7767459610736548, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.836310386657715, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8686792850494385, "num_tokens": 232985142.0, "step": 6106 }, { "epoch": 0.7768731713522452, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84356689453125, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8718733787536621, "num_tokens": 233020810.0, "step": 6107 }, { "epoch": 0.7770003816308357, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.808431625366211, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8543095588684082, "num_tokens": 233055674.0, "step": 6108 }, { "epoch": 0.7771275919094263, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.882769584655762, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8572483062744141, "num_tokens": 233084506.0, "step": 6109 }, { "epoch": 0.7772548021880168, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.795489311218262, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8789258599281311, "num_tokens": 233120172.0, "step": 6110 }, { "epoch": 0.7773820124666073, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.763720512390137, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8526033163070679, "num_tokens": 233160075.0, "step": 6111 }, { "epoch": 0.7775092227451978, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.817192077636719, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8542996644973755, "num_tokens": 233201714.0, "step": 6112 }, { "epoch": 0.7776364330237884, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.840688705444336, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8683048486709595, "num_tokens": 233237642.0, "step": 6113 }, { "epoch": 0.7777636433023788, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.824553489685059, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8673753142356873, "num_tokens": 233279712.0, "step": 6114 }, { "epoch": 0.7778908535809693, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.802501678466797, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8681240081787109, "num_tokens": 233322441.0, "step": 6115 }, { "epoch": 0.7780180638595598, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.816718101501465, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8557730913162231, "num_tokens": 233371053.0, "step": 6116 }, { "epoch": 0.7781452741381504, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.860527992248535, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.863381564617157, "num_tokens": 233405372.0, "step": 6117 }, { "epoch": 0.7782724844167409, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.815024375915527, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8691471815109253, "num_tokens": 233447209.0, "step": 6118 }, { "epoch": 0.7783996946953314, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.852132797241211, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8644245862960815, "num_tokens": 233484621.0, "step": 6119 }, { "epoch": 0.7785269049739219, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.871878623962402, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8452288508415222, "num_tokens": 233520756.0, "step": 6120 }, { "epoch": 0.7786541152525124, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.830696105957031, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8544949889183044, "num_tokens": 233559120.0, "step": 6121 }, { "epoch": 0.7787813255311029, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.852225303649902, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8736401796340942, "num_tokens": 233590513.0, "step": 6122 }, { "epoch": 0.7789085358096934, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.809081077575684, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8781505227088928, "num_tokens": 233625986.0, "step": 6123 }, { "epoch": 0.779035746088284, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.829931259155273, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8716254830360413, "num_tokens": 233655097.0, "step": 6124 }, { "epoch": 0.7791629563668745, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.836054801940918, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8524283766746521, "num_tokens": 233687815.0, "step": 6125 }, { "epoch": 0.7792901666454649, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.84188175201416, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8753658533096313, "num_tokens": 233722611.0, "step": 6126 }, { "epoch": 0.7794173769240554, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.83210563659668, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8501632213592529, "num_tokens": 233765345.0, "step": 6127 }, { "epoch": 0.779544587202646, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.816802024841309, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8644688725471497, "num_tokens": 233804358.0, "step": 6128 }, { "epoch": 0.7796717974812365, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.82242202758789, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8512558341026306, "num_tokens": 233845117.0, "step": 6129 }, { "epoch": 0.779799007759827, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.802956581115723, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8553382158279419, "num_tokens": 233884091.0, "step": 6130 }, { "epoch": 0.7799262180384176, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.871272087097168, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8600564002990723, "num_tokens": 233922590.0, "step": 6131 }, { "epoch": 0.780053428317008, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.844728469848633, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8653202652931213, "num_tokens": 233954138.0, "step": 6132 }, { "epoch": 0.7801806385955985, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.864139556884766, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8718920946121216, "num_tokens": 233988035.0, "step": 6133 }, { "epoch": 0.780307848874189, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.81912899017334, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8525843620300293, "num_tokens": 234023747.0, "step": 6134 }, { "epoch": 0.7804350591527796, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.794981956481934, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8708943724632263, "num_tokens": 234064182.0, "step": 6135 }, { "epoch": 0.7805622694313701, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.827288627624512, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8455865979194641, "num_tokens": 234106533.0, "step": 6136 }, { "epoch": 0.7806894797099606, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.794445991516113, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8748816251754761, "num_tokens": 234145492.0, "step": 6137 }, { "epoch": 0.780816689988551, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.867537498474121, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8648097515106201, "num_tokens": 234181709.0, "step": 6138 }, { "epoch": 0.7809439002671416, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.845974922180176, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8630595207214355, "num_tokens": 234217854.0, "step": 6139 }, { "epoch": 0.7810711105457321, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.826788902282715, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8617671132087708, "num_tokens": 234255769.0, "step": 6140 }, { "epoch": 0.7811983208243226, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.848673820495605, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8493907451629639, "num_tokens": 234290828.0, "step": 6141 }, { "epoch": 0.7813255311029131, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.874671936035156, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8624242544174194, "num_tokens": 234326336.0, "step": 6142 }, { "epoch": 0.7814527413815037, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.830748558044434, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8566880226135254, "num_tokens": 234366210.0, "step": 6143 }, { "epoch": 0.7815799516600941, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.877412796020508, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8536967635154724, "num_tokens": 234398975.0, "step": 6144 }, { "epoch": 0.7817071619386846, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.863105773925781, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.856971800327301, "num_tokens": 234437746.0, "step": 6145 }, { "epoch": 0.7818343722172751, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.86236572265625, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8680882453918457, "num_tokens": 234478402.0, "step": 6146 }, { "epoch": 0.7819615824958657, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.907828330993652, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8478933572769165, "num_tokens": 234509699.0, "step": 6147 }, { "epoch": 0.7820887927744562, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.814467430114746, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8574923276901245, "num_tokens": 234551515.0, "step": 6148 }, { "epoch": 0.7822160030530467, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.893345832824707, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8695360422134399, "num_tokens": 234582811.0, "step": 6149 }, { "epoch": 0.7823432133316371, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.82858943939209, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8547980785369873, "num_tokens": 234625299.0, "step": 6150 }, { "epoch": 0.7824704236102277, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.825560569763184, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8687126636505127, "num_tokens": 234665641.0, "step": 6151 }, { "epoch": 0.7825976338888182, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.884306907653809, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8619680404663086, "num_tokens": 234704668.0, "step": 6152 }, { "epoch": 0.7827248441674087, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.896516799926758, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8668885231018066, "num_tokens": 234738080.0, "step": 6153 }, { "epoch": 0.7828520544459993, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.910338401794434, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8601779937744141, "num_tokens": 234781077.0, "step": 6154 }, { "epoch": 0.7829792647245898, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.888100624084473, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8466876745223999, "num_tokens": 234809374.0, "step": 6155 }, { "epoch": 0.7831064750031802, "ewc_loss": 0.021728515625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.887919425964355, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8505977988243103, "num_tokens": 234845585.0, "step": 6156 }, { "epoch": 0.7832336852817707, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.85915470123291, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8609840869903564, "num_tokens": 234883891.0, "step": 6157 }, { "epoch": 0.7833608955603613, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.827593803405762, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8610312938690186, "num_tokens": 234919445.0, "step": 6158 }, { "epoch": 0.7834881058389518, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.881485939025879, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8646838665008545, "num_tokens": 234951465.0, "step": 6159 }, { "epoch": 0.7836153161175423, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.89164924621582, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8538823127746582, "num_tokens": 234989970.0, "step": 6160 }, { "epoch": 0.7837425263961328, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.835768699645996, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8636980056762695, "num_tokens": 235029673.0, "step": 6161 }, { "epoch": 0.7838697366747234, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.890636444091797, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8617249131202698, "num_tokens": 235065389.0, "step": 6162 }, { "epoch": 0.7839969469533138, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.88186264038086, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8734592199325562, "num_tokens": 235099795.0, "step": 6163 }, { "epoch": 0.7841241572319043, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.829266548156738, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8630523085594177, "num_tokens": 235137475.0, "step": 6164 }, { "epoch": 0.7842513675104948, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.87684154510498, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8576366901397705, "num_tokens": 235178454.0, "step": 6165 }, { "epoch": 0.7843785777890854, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.823494911193848, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8800280094146729, "num_tokens": 235216139.0, "step": 6166 }, { "epoch": 0.7845057880676759, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.827276229858398, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8772264719009399, "num_tokens": 235253571.0, "step": 6167 }, { "epoch": 0.7846329983462664, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.867602348327637, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8463168740272522, "num_tokens": 235293950.0, "step": 6168 }, { "epoch": 0.7847602086248568, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.88036823272705, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8664844632148743, "num_tokens": 235336419.0, "step": 6169 }, { "epoch": 0.7848874189034474, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.852928161621094, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8673003911972046, "num_tokens": 235376039.0, "step": 6170 }, { "epoch": 0.7850146291820379, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.897723197937012, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.850838303565979, "num_tokens": 235412736.0, "step": 6171 }, { "epoch": 0.7851418394606284, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.838356018066406, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8621717691421509, "num_tokens": 235457318.0, "step": 6172 }, { "epoch": 0.785269049739219, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.859180450439453, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8522124290466309, "num_tokens": 235491471.0, "step": 6173 }, { "epoch": 0.7853962600178095, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.84045124053955, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8726993799209595, "num_tokens": 235531324.0, "step": 6174 }, { "epoch": 0.7855234702963999, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.875516891479492, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8493070006370544, "num_tokens": 235573233.0, "step": 6175 }, { "epoch": 0.7856506805749904, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.811380386352539, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8629825711250305, "num_tokens": 235614400.0, "step": 6176 }, { "epoch": 0.785777890853581, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.850774765014648, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8691064119338989, "num_tokens": 235652247.0, "step": 6177 }, { "epoch": 0.7859051011321715, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.891377449035645, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8588330745697021, "num_tokens": 235681810.0, "step": 6178 }, { "epoch": 0.786032311410762, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.833405494689941, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8713852763175964, "num_tokens": 235718422.0, "step": 6179 }, { "epoch": 0.7861595216893525, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.856688499450684, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8571128845214844, "num_tokens": 235758886.0, "step": 6180 }, { "epoch": 0.786286731967943, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.917081832885742, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8657180070877075, "num_tokens": 235792270.0, "step": 6181 }, { "epoch": 0.7864139422465335, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.86013126373291, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.852419376373291, "num_tokens": 235826678.0, "step": 6182 }, { "epoch": 0.786541152525124, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.911205291748047, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8736791610717773, "num_tokens": 235860437.0, "step": 6183 }, { "epoch": 0.7866683628037145, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.890485763549805, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8670724034309387, "num_tokens": 235900865.0, "step": 6184 }, { "epoch": 0.7867955730823051, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.851422309875488, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8471921682357788, "num_tokens": 235937419.0, "step": 6185 }, { "epoch": 0.7869227833608956, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.865915298461914, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8532461524009705, "num_tokens": 235976425.0, "step": 6186 }, { "epoch": 0.787049993639486, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.903995513916016, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8558955192565918, "num_tokens": 236012968.0, "step": 6187 }, { "epoch": 0.7871772039180766, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.858391761779785, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.863623857498169, "num_tokens": 236052462.0, "step": 6188 }, { "epoch": 0.7873044141966671, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.861860275268555, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8773435354232788, "num_tokens": 236087604.0, "step": 6189 }, { "epoch": 0.7874316244752576, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.846884727478027, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8625411987304688, "num_tokens": 236126344.0, "step": 6190 }, { "epoch": 0.7875588347538481, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.863017082214355, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8701004981994629, "num_tokens": 236161913.0, "step": 6191 }, { "epoch": 0.7876860450324387, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.89345645904541, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8657827377319336, "num_tokens": 236200612.0, "step": 6192 }, { "epoch": 0.7878132553110291, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.900989532470703, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8682010173797607, "num_tokens": 236238136.0, "step": 6193 }, { "epoch": 0.7879404655896196, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.845046997070312, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8588871955871582, "num_tokens": 236281303.0, "step": 6194 }, { "epoch": 0.7880676758682101, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.870616912841797, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8592685461044312, "num_tokens": 236322355.0, "step": 6195 }, { "epoch": 0.7881948861468007, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.830082893371582, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8610810041427612, "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": 10.853523254394531, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8634759783744812, "num_tokens": 236391581.0, "step": 6197 }, { "epoch": 0.7884493067039817, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.853114128112793, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8710078001022339, "num_tokens": 236426003.0, "step": 6198 }, { "epoch": 0.7885765169825721, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.848978042602539, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8724291324615479, "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": 10.857940673828125, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.873035192489624, "num_tokens": 236499826.0, "step": 6200 }, { "epoch": 0.7888309375397532, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.85890007019043, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8743738532066345, "num_tokens": 236538615.0, "step": 6201 }, { "epoch": 0.7889581478183437, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.865358352661133, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8603072762489319, "num_tokens": 236577860.0, "step": 6202 }, { "epoch": 0.7890853580969343, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.859869956970215, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8773536682128906, "num_tokens": 236619041.0, "step": 6203 }, { "epoch": 0.7892125683755248, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.892391204833984, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8654718995094299, "num_tokens": 236655788.0, "step": 6204 }, { "epoch": 0.7893397786541152, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.868050575256348, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8736953735351562, "num_tokens": 236691532.0, "step": 6205 }, { "epoch": 0.7894669889327057, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.844975471496582, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8692584037780762, "num_tokens": 236733792.0, "step": 6206 }, { "epoch": 0.7895941992112963, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.844030380249023, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8745986223220825, "num_tokens": 236775454.0, "step": 6207 }, { "epoch": 0.7897214094898868, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.885682106018066, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8731107711791992, "num_tokens": 236811572.0, "step": 6208 }, { "epoch": 0.7898486197684773, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.868830680847168, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8472949266433716, "num_tokens": 236851018.0, "step": 6209 }, { "epoch": 0.7899758300470678, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.843707084655762, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8672025799751282, "num_tokens": 236891871.0, "step": 6210 }, { "epoch": 0.7901030403256584, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.917585372924805, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8662191033363342, "num_tokens": 236926771.0, "step": 6211 }, { "epoch": 0.7902302506042488, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.834074020385742, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8504458069801331, "num_tokens": 236964768.0, "step": 6212 }, { "epoch": 0.7903574608828393, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.857608795166016, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8513590097427368, "num_tokens": 237005075.0, "step": 6213 }, { "epoch": 0.7904846711614298, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.910168647766113, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.86467444896698, "num_tokens": 237035517.0, "step": 6214 }, { "epoch": 0.7906118814400204, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.863327980041504, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8797874450683594, "num_tokens": 237076319.0, "step": 6215 }, { "epoch": 0.7907390917186109, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.807608604431152, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.853449821472168, "num_tokens": 237117935.0, "step": 6216 }, { "epoch": 0.7908663019972014, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.88469123840332, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8723607063293457, "num_tokens": 237155789.0, "step": 6217 }, { "epoch": 0.7909935122757918, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.827899932861328, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8683955669403076, "num_tokens": 237198870.0, "step": 6218 }, { "epoch": 0.7911207225543824, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.891646385192871, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8680097460746765, "num_tokens": 237243084.0, "step": 6219 }, { "epoch": 0.7912479328329729, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.898548126220703, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8561409711837769, "num_tokens": 237277245.0, "step": 6220 }, { "epoch": 0.7913751431115634, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.894271850585938, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8676896095275879, "num_tokens": 237317841.0, "step": 6221 }, { "epoch": 0.791502353390154, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.93881607055664, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.852800726890564, "num_tokens": 237349327.0, "step": 6222 }, { "epoch": 0.7916295636687445, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.848002433776855, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8486917018890381, "num_tokens": 237396179.0, "step": 6223 }, { "epoch": 0.7917567739473349, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.829512596130371, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.860090970993042, "num_tokens": 237437766.0, "step": 6224 }, { "epoch": 0.7918839842259254, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.834122657775879, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8678834438323975, "num_tokens": 237479239.0, "step": 6225 }, { "epoch": 0.792011194504516, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.860233306884766, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8773597478866577, "num_tokens": 237514641.0, "step": 6226 }, { "epoch": 0.7921384047831065, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.852190017700195, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8735872507095337, "num_tokens": 237555319.0, "step": 6227 }, { "epoch": 0.792265615061697, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.874723434448242, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8656973838806152, "num_tokens": 237597786.0, "step": 6228 }, { "epoch": 0.7923928253402875, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.834490776062012, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8743621706962585, "num_tokens": 237636584.0, "step": 6229 }, { "epoch": 0.792520035618878, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.870626449584961, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8546837568283081, "num_tokens": 237676000.0, "step": 6230 }, { "epoch": 0.7926472458974685, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.836856842041016, "learning_rate": 1e-06, "loss": 0.3455, "mean_token_accuracy": 0.8895902633666992, "num_tokens": 237713003.0, "step": 6231 }, { "epoch": 0.792774456176059, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.8875732421875, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8650513291358948, "num_tokens": 237742054.0, "step": 6232 }, { "epoch": 0.7929016664546495, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.836593627929688, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8558797836303711, "num_tokens": 237783147.0, "step": 6233 }, { "epoch": 0.7930288767332401, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.903621673583984, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8577748537063599, "num_tokens": 237814322.0, "step": 6234 }, { "epoch": 0.7931560870118306, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.88203239440918, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8694121837615967, "num_tokens": 237849278.0, "step": 6235 }, { "epoch": 0.793283297290421, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.825407028198242, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8672358393669128, "num_tokens": 237893434.0, "step": 6236 }, { "epoch": 0.7934105075690115, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.838765144348145, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8713216185569763, "num_tokens": 237932135.0, "step": 6237 }, { "epoch": 0.7935377178476021, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.900749206542969, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8583202362060547, "num_tokens": 237966320.0, "step": 6238 }, { "epoch": 0.7936649281261926, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.859535217285156, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8677124977111816, "num_tokens": 238006365.0, "step": 6239 }, { "epoch": 0.7937921384047831, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.855530738830566, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8616727590560913, "num_tokens": 238039307.0, "step": 6240 }, { "epoch": 0.7939193486833737, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.878244400024414, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8495854139328003, "num_tokens": 238078015.0, "step": 6241 }, { "epoch": 0.7940465589619641, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.0558319091796875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.875734329223633, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.860801100730896, "num_tokens": 238116790.0, "step": 6242 }, { "epoch": 0.7941737692405546, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.887151718139648, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8698552250862122, "num_tokens": 238155330.0, "step": 6243 }, { "epoch": 0.7943009795191451, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.890914916992188, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8479125499725342, "num_tokens": 238194492.0, "step": 6244 }, { "epoch": 0.7944281897977357, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.856510162353516, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.864000678062439, "num_tokens": 238230451.0, "step": 6245 }, { "epoch": 0.7945554000763262, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.900910377502441, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8709902167320251, "num_tokens": 238266331.0, "step": 6246 }, { "epoch": 0.7946826103549167, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.864008903503418, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8525031805038452, "num_tokens": 238307753.0, "step": 6247 }, { "epoch": 0.7948098206335071, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.863866806030273, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8639930486679077, "num_tokens": 238349644.0, "step": 6248 }, { "epoch": 0.7949370309120977, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.890992164611816, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8498830795288086, "num_tokens": 238386308.0, "step": 6249 }, { "epoch": 0.7950642411906882, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.862730026245117, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8484908938407898, "num_tokens": 238427935.0, "step": 6250 }, { "epoch": 0.7951914514692787, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.880599021911621, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8686434626579285, "num_tokens": 238463639.0, "step": 6251 }, { "epoch": 0.7953186617478692, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.876187324523926, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8765546083450317, "num_tokens": 238498295.0, "step": 6252 }, { "epoch": 0.7954458720264598, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.848468780517578, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8634530305862427, "num_tokens": 238538109.0, "step": 6253 }, { "epoch": 0.7955730823050502, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.913152694702148, "learning_rate": 1e-06, "loss": 0.4679, "mean_token_accuracy": 0.8497910499572754, "num_tokens": 238574415.0, "step": 6254 }, { "epoch": 0.7957002925836407, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.858853340148926, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8647042512893677, "num_tokens": 238616368.0, "step": 6255 }, { "epoch": 0.7958275028622313, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.872042655944824, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8671667575836182, "num_tokens": 238650634.0, "step": 6256 }, { "epoch": 0.7959547131408218, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.901659965515137, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8581960797309875, "num_tokens": 238684383.0, "step": 6257 }, { "epoch": 0.7960819234194123, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.878588676452637, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8605626225471497, "num_tokens": 238722863.0, "step": 6258 }, { "epoch": 0.7962091336980028, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.846631050109863, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8644489645957947, "num_tokens": 238765027.0, "step": 6259 }, { "epoch": 0.7963363439765934, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.915820121765137, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8678644895553589, "num_tokens": 238801950.0, "step": 6260 }, { "epoch": 0.7964635542551838, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.898616790771484, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8538436889648438, "num_tokens": 238840963.0, "step": 6261 }, { "epoch": 0.7965907645337743, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.854416847229004, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.850472092628479, "num_tokens": 238876167.0, "step": 6262 }, { "epoch": 0.7967179748123648, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.852445602416992, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8695957660675049, "num_tokens": 238910293.0, "step": 6263 }, { "epoch": 0.7968451850909554, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.880417823791504, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8608320951461792, "num_tokens": 238947279.0, "step": 6264 }, { "epoch": 0.7969723953695459, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.894990921020508, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8519541621208191, "num_tokens": 238984148.0, "step": 6265 }, { "epoch": 0.7970996056481364, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.90274715423584, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8494447469711304, "num_tokens": 239022370.0, "step": 6266 }, { "epoch": 0.7972268159267268, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.866310119628906, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8595916032791138, "num_tokens": 239064542.0, "step": 6267 }, { "epoch": 0.7973540262053174, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.894876480102539, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8615906238555908, "num_tokens": 239103193.0, "step": 6268 }, { "epoch": 0.7974812364839079, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.8944673538208, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8701760172843933, "num_tokens": 239137952.0, "step": 6269 }, { "epoch": 0.7976084467624984, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.924595832824707, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8538598418235779, "num_tokens": 239174600.0, "step": 6270 }, { "epoch": 0.797735657041089, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.88505744934082, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.867569088935852, "num_tokens": 239206157.0, "step": 6271 }, { "epoch": 0.7978628673196795, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.875814437866211, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8678854703903198, "num_tokens": 239243094.0, "step": 6272 }, { "epoch": 0.7979900775982699, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.833541870117188, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8639264106750488, "num_tokens": 239281208.0, "step": 6273 }, { "epoch": 0.7981172878768604, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.88650131225586, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8588112592697144, "num_tokens": 239320159.0, "step": 6274 }, { "epoch": 0.798244498155451, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.853271484375, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8513702750205994, "num_tokens": 239360160.0, "step": 6275 }, { "epoch": 0.7983717084340415, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.870285034179688, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8650699853897095, "num_tokens": 239398101.0, "step": 6276 }, { "epoch": 0.798498918712632, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.854558944702148, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8784428238868713, "num_tokens": 239438560.0, "step": 6277 }, { "epoch": 0.7986261289912225, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.890284538269043, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8568090200424194, "num_tokens": 239475851.0, "step": 6278 }, { "epoch": 0.798753339269813, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.845577239990234, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8680206537246704, "num_tokens": 239519165.0, "step": 6279 }, { "epoch": 0.7988805495484035, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.89044189453125, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8509984016418457, "num_tokens": 239560602.0, "step": 6280 }, { "epoch": 0.799007759826994, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.910792350769043, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.864250898361206, "num_tokens": 239591933.0, "step": 6281 }, { "epoch": 0.7991349701055845, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.956870079040527, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8571852445602417, "num_tokens": 239625495.0, "step": 6282 }, { "epoch": 0.7992621803841751, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.930353164672852, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.878160834312439, "num_tokens": 239658249.0, "step": 6283 }, { "epoch": 0.7993893906627656, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.928312301635742, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8539945483207703, "num_tokens": 239690194.0, "step": 6284 }, { "epoch": 0.799516600941356, "ewc_loss": 0.021728515625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 10.860431671142578, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8606101274490356, "num_tokens": 239728315.0, "step": 6285 }, { "epoch": 0.7996438112199465, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.853707313537598, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8678669929504395, "num_tokens": 239764505.0, "step": 6286 }, { "epoch": 0.7997710214985371, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.922566413879395, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8477898240089417, "num_tokens": 239809461.0, "step": 6287 }, { "epoch": 0.7998982317771276, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.884330749511719, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8575330972671509, "num_tokens": 239845232.0, "step": 6288 }, { "epoch": 0.8000254420557181, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.876358032226562, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8688665628433228, "num_tokens": 239885489.0, "step": 6289 }, { "epoch": 0.8001526523343087, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.8786039352417, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8504873514175415, "num_tokens": 239921159.0, "step": 6290 }, { "epoch": 0.8002798626128991, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.915270805358887, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8671522736549377, "num_tokens": 239960879.0, "step": 6291 }, { "epoch": 0.8004070728914896, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.852667808532715, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8520674705505371, "num_tokens": 239994302.0, "step": 6292 }, { "epoch": 0.8005342831700801, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.879729270935059, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8696849346160889, "num_tokens": 240027759.0, "step": 6293 }, { "epoch": 0.8006614934486707, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.885643005371094, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8577759861946106, "num_tokens": 240065049.0, "step": 6294 }, { "epoch": 0.8007887037272612, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.95091724395752, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.8485561609268188, "num_tokens": 240101453.0, "step": 6295 }, { "epoch": 0.8009159140058517, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.876822471618652, "learning_rate": 1e-06, "loss": 0.478, "mean_token_accuracy": 0.8411770462989807, "num_tokens": 240136831.0, "step": 6296 }, { "epoch": 0.8010431242844421, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.913431167602539, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8613156080245972, "num_tokens": 240173364.0, "step": 6297 }, { "epoch": 0.8011703345630327, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.937124252319336, "learning_rate": 1e-06, "loss": 0.4783, "mean_token_accuracy": 0.840976357460022, "num_tokens": 240206718.0, "step": 6298 }, { "epoch": 0.8012975448416232, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.885895729064941, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8499168157577515, "num_tokens": 240245633.0, "step": 6299 }, { "epoch": 0.8014247551202137, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.950782775878906, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8766292333602905, "num_tokens": 240283887.0, "step": 6300 }, { "epoch": 0.8015519653988042, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.907218933105469, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.854185163974762, "num_tokens": 240322208.0, "step": 6301 }, { "epoch": 0.8016791756773948, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.931510925292969, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.866068422794342, "num_tokens": 240356154.0, "step": 6302 }, { "epoch": 0.8018063859559852, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.929970741271973, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8609734773635864, "num_tokens": 240389103.0, "step": 6303 }, { "epoch": 0.8019335962345757, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.913736343383789, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.879225492477417, "num_tokens": 240429545.0, "step": 6304 }, { "epoch": 0.8020608065131662, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.872503280639648, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8791338205337524, "num_tokens": 240468556.0, "step": 6305 }, { "epoch": 0.8021880167917568, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.912101745605469, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8633939623832703, "num_tokens": 240508534.0, "step": 6306 }, { "epoch": 0.8023152270703473, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.940262794494629, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8619908094406128, "num_tokens": 240540833.0, "step": 6307 }, { "epoch": 0.8024424373489378, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.929842948913574, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8544900417327881, "num_tokens": 240577768.0, "step": 6308 }, { "epoch": 0.8025696476275284, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.91865348815918, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8582538366317749, "num_tokens": 240610607.0, "step": 6309 }, { "epoch": 0.8026968579061188, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.904889106750488, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8590019941329956, "num_tokens": 240646056.0, "step": 6310 }, { "epoch": 0.8028240681847093, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.910322189331055, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8692899942398071, "num_tokens": 240682513.0, "step": 6311 }, { "epoch": 0.8029512784632998, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.929354667663574, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8632495403289795, "num_tokens": 240716311.0, "step": 6312 }, { "epoch": 0.8030784887418904, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.967244148254395, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8542759418487549, "num_tokens": 240758873.0, "step": 6313 }, { "epoch": 0.8032056990204809, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.919075012207031, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8676337599754333, "num_tokens": 240797163.0, "step": 6314 }, { "epoch": 0.8033329092990714, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.938827514648438, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8734384775161743, "num_tokens": 240835719.0, "step": 6315 }, { "epoch": 0.8034601195776618, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.911038398742676, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8558762073516846, "num_tokens": 240874359.0, "step": 6316 }, { "epoch": 0.8035873298562524, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.896726608276367, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8754485249519348, "num_tokens": 240909125.0, "step": 6317 }, { "epoch": 0.8037145401348429, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.926158905029297, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8671414852142334, "num_tokens": 240949391.0, "step": 6318 }, { "epoch": 0.8038417504134334, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.911884307861328, "learning_rate": 1e-06, "loss": 0.4696, "mean_token_accuracy": 0.8525992035865784, "num_tokens": 240987390.0, "step": 6319 }, { "epoch": 0.803968960692024, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.926809310913086, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8555571436882019, "num_tokens": 241025359.0, "step": 6320 }, { "epoch": 0.8040961709706145, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.928560256958008, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8570247888565063, "num_tokens": 241060532.0, "step": 6321 }, { "epoch": 0.8042233812492049, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.89676570892334, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.869006335735321, "num_tokens": 241098547.0, "step": 6322 }, { "epoch": 0.8043505915277954, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.922513961791992, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8623859882354736, "num_tokens": 241131091.0, "step": 6323 }, { "epoch": 0.804477801806386, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.976569175720215, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.851986289024353, "num_tokens": 241164495.0, "step": 6324 }, { "epoch": 0.8046050120849765, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.903000831604004, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8561708927154541, "num_tokens": 241203777.0, "step": 6325 }, { "epoch": 0.804732222363567, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.896760940551758, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8559211492538452, "num_tokens": 241243219.0, "step": 6326 }, { "epoch": 0.8048594326421575, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.88632869720459, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8461306095123291, "num_tokens": 241284547.0, "step": 6327 }, { "epoch": 0.804986642920748, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.839860916137695, "learning_rate": 1e-06, "loss": 0.495, "mean_token_accuracy": 0.845424234867096, "num_tokens": 241331110.0, "step": 6328 }, { "epoch": 0.8051138531993385, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.907805442810059, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8615408539772034, "num_tokens": 241372720.0, "step": 6329 }, { "epoch": 0.805241063477929, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.87148380279541, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.863773763179779, "num_tokens": 241408173.0, "step": 6330 }, { "epoch": 0.8053682737565195, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.950345039367676, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8682963252067566, "num_tokens": 241445021.0, "step": 6331 }, { "epoch": 0.8054954840351101, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.885456085205078, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.866770327091217, "num_tokens": 241481874.0, "step": 6332 }, { "epoch": 0.8056226943137006, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.906394958496094, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8468086123466492, "num_tokens": 241523140.0, "step": 6333 }, { "epoch": 0.805749904592291, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.878872871398926, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8446212410926819, "num_tokens": 241563424.0, "step": 6334 }, { "epoch": 0.8058771148708815, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.870548248291016, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8747431039810181, "num_tokens": 241600004.0, "step": 6335 }, { "epoch": 0.8060043251494721, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.926220893859863, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8609136343002319, "num_tokens": 241642704.0, "step": 6336 }, { "epoch": 0.8061315354280626, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.87548542022705, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8673669099807739, "num_tokens": 241679176.0, "step": 6337 }, { "epoch": 0.8062587457066531, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.897566795349121, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8667458891868591, "num_tokens": 241721168.0, "step": 6338 }, { "epoch": 0.8063859559852437, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.895577430725098, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8755881786346436, "num_tokens": 241759735.0, "step": 6339 }, { "epoch": 0.8065131662638341, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.893911361694336, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8623439073562622, "num_tokens": 241799445.0, "step": 6340 }, { "epoch": 0.8066403765424246, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.90539836883545, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8662213087081909, "num_tokens": 241839333.0, "step": 6341 }, { "epoch": 0.8067675868210151, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.89997386932373, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8503579497337341, "num_tokens": 241880248.0, "step": 6342 }, { "epoch": 0.8068947970996057, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.885807991027832, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8752419948577881, "num_tokens": 241926717.0, "step": 6343 }, { "epoch": 0.8070220073781962, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.870615005493164, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8523626923561096, "num_tokens": 241966080.0, "step": 6344 }, { "epoch": 0.8071492176567867, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.89555549621582, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8590680360794067, "num_tokens": 242011975.0, "step": 6345 }, { "epoch": 0.8072764279353771, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.87875747680664, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8760676383972168, "num_tokens": 242046320.0, "step": 6346 }, { "epoch": 0.8074036382139677, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.896712303161621, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8637582659721375, "num_tokens": 242085157.0, "step": 6347 }, { "epoch": 0.8075308484925582, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.9393949508667, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8581975698471069, "num_tokens": 242116548.0, "step": 6348 }, { "epoch": 0.8076580587711487, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.93315601348877, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8684893846511841, "num_tokens": 242151883.0, "step": 6349 }, { "epoch": 0.8077852690497392, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.90019702911377, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8596082329750061, "num_tokens": 242190486.0, "step": 6350 }, { "epoch": 0.8079124793283298, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.911261558532715, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8659061789512634, "num_tokens": 242232720.0, "step": 6351 }, { "epoch": 0.8080396896069202, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.905987739562988, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8633861541748047, "num_tokens": 242273926.0, "step": 6352 }, { "epoch": 0.8081668998855107, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.874628067016602, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8513555526733398, "num_tokens": 242313484.0, "step": 6353 }, { "epoch": 0.8082941101641012, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.899903297424316, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8560025095939636, "num_tokens": 242352101.0, "step": 6354 }, { "epoch": 0.8084213204426918, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.980419158935547, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8776015043258667, "num_tokens": 242392958.0, "step": 6355 }, { "epoch": 0.8085485307212823, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.862218856811523, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8627874851226807, "num_tokens": 242432327.0, "step": 6356 }, { "epoch": 0.8086757409998728, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.865910530090332, "learning_rate": 1e-06, "loss": 0.4887, "mean_token_accuracy": 0.8410764932632446, "num_tokens": 242473714.0, "step": 6357 }, { "epoch": 0.8088029512784632, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.944426536560059, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8695145845413208, "num_tokens": 242513179.0, "step": 6358 }, { "epoch": 0.8089301615570538, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.872177124023438, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8631355166435242, "num_tokens": 242549398.0, "step": 6359 }, { "epoch": 0.8090573718356443, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.921895980834961, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8564649820327759, "num_tokens": 242583705.0, "step": 6360 }, { "epoch": 0.8091845821142348, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.87112045288086, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8665314316749573, "num_tokens": 242625900.0, "step": 6361 }, { "epoch": 0.8093117923928254, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.876082420349121, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8504593968391418, "num_tokens": 242665766.0, "step": 6362 }, { "epoch": 0.8094390026714159, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.879916191101074, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8655012845993042, "num_tokens": 242707634.0, "step": 6363 }, { "epoch": 0.8095662129500064, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.861742973327637, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8573285341262817, "num_tokens": 242746791.0, "step": 6364 }, { "epoch": 0.8096934232285968, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.883681297302246, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8757786750793457, "num_tokens": 242784861.0, "step": 6365 }, { "epoch": 0.8098206335071874, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.845041275024414, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8765873312950134, "num_tokens": 242817612.0, "step": 6366 }, { "epoch": 0.8099478437857779, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.905391693115234, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.869093656539917, "num_tokens": 242854951.0, "step": 6367 }, { "epoch": 0.8100750540643684, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.828056335449219, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8771567344665527, "num_tokens": 242894166.0, "step": 6368 }, { "epoch": 0.8102022643429589, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.905387878417969, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8675353527069092, "num_tokens": 242936230.0, "step": 6369 }, { "epoch": 0.8103294746215495, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.925244331359863, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.873598575592041, "num_tokens": 242973148.0, "step": 6370 }, { "epoch": 0.8104566849001399, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.882614135742188, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8635509014129639, "num_tokens": 243013029.0, "step": 6371 }, { "epoch": 0.8105838951787304, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.876643180847168, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8610255718231201, "num_tokens": 243054847.0, "step": 6372 }, { "epoch": 0.810711105457321, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.894392013549805, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8556126356124878, "num_tokens": 243096524.0, "step": 6373 }, { "epoch": 0.8108383157359115, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.917057991027832, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8553069829940796, "num_tokens": 243132397.0, "step": 6374 }, { "epoch": 0.810965526014502, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.90438175201416, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.849207878112793, "num_tokens": 243171790.0, "step": 6375 }, { "epoch": 0.8110927362930925, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.949967384338379, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.861417293548584, "num_tokens": 243208217.0, "step": 6376 }, { "epoch": 0.811219946571683, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.941384315490723, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8669317960739136, "num_tokens": 243245260.0, "step": 6377 }, { "epoch": 0.8113471568502735, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.885107040405273, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8708755373954773, "num_tokens": 243283924.0, "step": 6378 }, { "epoch": 0.811474367128864, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.928898811340332, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8703962564468384, "num_tokens": 243321977.0, "step": 6379 }, { "epoch": 0.8116015774074545, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.874068260192871, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.858704149723053, "num_tokens": 243359349.0, "step": 6380 }, { "epoch": 0.8117287876860451, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.89130687713623, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.865024209022522, "num_tokens": 243395722.0, "step": 6381 }, { "epoch": 0.8118559979646356, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.95111083984375, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8523051738739014, "num_tokens": 243429022.0, "step": 6382 }, { "epoch": 0.811983208243226, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.879449844360352, "learning_rate": 1e-06, "loss": 0.4925, "mean_token_accuracy": 0.8417802453041077, "num_tokens": 243468474.0, "step": 6383 }, { "epoch": 0.8121104185218165, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.940184593200684, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.878178596496582, "num_tokens": 243503045.0, "step": 6384 }, { "epoch": 0.8122376288004071, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.904662132263184, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8518110513687134, "num_tokens": 243543865.0, "step": 6385 }, { "epoch": 0.8123648390789976, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.926095008850098, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8561339378356934, "num_tokens": 243581258.0, "step": 6386 }, { "epoch": 0.8124920493575881, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.927454948425293, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8571339845657349, "num_tokens": 243622777.0, "step": 6387 }, { "epoch": 0.8126192596361786, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.921475410461426, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8504788279533386, "num_tokens": 243664549.0, "step": 6388 }, { "epoch": 0.8127464699147691, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.933453559875488, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8566330671310425, "num_tokens": 243710408.0, "step": 6389 }, { "epoch": 0.8128736801933596, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.861776351928711, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8632692098617554, "num_tokens": 243752598.0, "step": 6390 }, { "epoch": 0.8130008904719501, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.904085159301758, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8590488433837891, "num_tokens": 243789805.0, "step": 6391 }, { "epoch": 0.8131281007505406, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.898343086242676, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8722445964813232, "num_tokens": 243824124.0, "step": 6392 }, { "epoch": 0.8132553110291312, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.916584968566895, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8621079325675964, "num_tokens": 243863448.0, "step": 6393 }, { "epoch": 0.8133825213077217, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.90072250366211, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8659143447875977, "num_tokens": 243902085.0, "step": 6394 }, { "epoch": 0.8135097315863121, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.90064811706543, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8570635318756104, "num_tokens": 243945349.0, "step": 6395 }, { "epoch": 0.8136369418649027, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.920000076293945, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.860977828502655, "num_tokens": 243982221.0, "step": 6396 }, { "epoch": 0.8137641521434932, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.96098518371582, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8493601083755493, "num_tokens": 244012576.0, "step": 6397 }, { "epoch": 0.8138913624220837, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.937105178833008, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8690981268882751, "num_tokens": 244052089.0, "step": 6398 }, { "epoch": 0.8140185727006742, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.912946701049805, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8637588024139404, "num_tokens": 244088712.0, "step": 6399 }, { "epoch": 0.8141457829792648, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.962745666503906, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8635290265083313, "num_tokens": 244127655.0, "step": 6400 }, { "epoch": 0.8142729932578552, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.965274810791016, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8671046495437622, "num_tokens": 244158588.0, "step": 6401 }, { "epoch": 0.8144002035364457, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.914576530456543, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8806543350219727, "num_tokens": 244193872.0, "step": 6402 }, { "epoch": 0.8145274138150362, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.948444366455078, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.847750186920166, "num_tokens": 244234055.0, "step": 6403 }, { "epoch": 0.8146546240936268, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.918456077575684, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8633540868759155, "num_tokens": 244275163.0, "step": 6404 }, { "epoch": 0.8147818343722173, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.906332969665527, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8593941926956177, "num_tokens": 244315241.0, "step": 6405 }, { "epoch": 0.8149090446508078, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.941576957702637, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8766664266586304, "num_tokens": 244351947.0, "step": 6406 }, { "epoch": 0.8150362549293982, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.916618347167969, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8598095178604126, "num_tokens": 244387706.0, "step": 6407 }, { "epoch": 0.8151634652079888, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.938613891601562, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8704043626785278, "num_tokens": 244428447.0, "step": 6408 }, { "epoch": 0.8152906754865793, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.982090950012207, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8474869728088379, "num_tokens": 244471191.0, "step": 6409 }, { "epoch": 0.8154178857651698, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.958428382873535, "learning_rate": 1e-06, "loss": 0.4993, "mean_token_accuracy": 0.8445115685462952, "num_tokens": 244509198.0, "step": 6410 }, { "epoch": 0.8155450960437604, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.974843978881836, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8541296124458313, "num_tokens": 244549677.0, "step": 6411 }, { "epoch": 0.8156723063223509, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00008487701416, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.850002110004425, "num_tokens": 244583586.0, "step": 6412 }, { "epoch": 0.8157995166009414, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.875615119934082, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8702342510223389, "num_tokens": 244621220.0, "step": 6413 }, { "epoch": 0.8159267268795318, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.996341705322266, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8694879412651062, "num_tokens": 244655941.0, "step": 6414 }, { "epoch": 0.8160539371581224, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.944302558898926, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8791674971580505, "num_tokens": 244685807.0, "step": 6415 }, { "epoch": 0.8161811474367129, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.926398277282715, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8595271706581116, "num_tokens": 244726866.0, "step": 6416 }, { "epoch": 0.8163083577153034, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.95042610168457, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8540428280830383, "num_tokens": 244767273.0, "step": 6417 }, { "epoch": 0.8164355679938939, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.956375122070312, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8575001955032349, "num_tokens": 244806111.0, "step": 6418 }, { "epoch": 0.8165627782724845, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.914590835571289, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8730023503303528, "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": 10.886222839355469, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8647018074989319, "num_tokens": 244883908.0, "step": 6420 }, { "epoch": 0.8168171988296654, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.954395294189453, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8542572855949402, "num_tokens": 244919090.0, "step": 6421 }, { "epoch": 0.8169444091082559, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.882140159606934, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8762500286102295, "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": 10.91768741607666, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8466945886611938, "num_tokens": 245004503.0, "step": 6423 }, { "epoch": 0.817198829665437, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.953304290771484, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8614980578422546, "num_tokens": 245045042.0, "step": 6424 }, { "epoch": 0.8173260399440275, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.905949592590332, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8604265451431274, "num_tokens": 245085416.0, "step": 6425 }, { "epoch": 0.8174532502226179, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.963753700256348, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.852534294128418, "num_tokens": 245116137.0, "step": 6426 }, { "epoch": 0.8175804605012085, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.870723724365234, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8742992877960205, "num_tokens": 245159181.0, "step": 6427 }, { "epoch": 0.817707670779799, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.893000602722168, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8533644080162048, "num_tokens": 245205199.0, "step": 6428 }, { "epoch": 0.8178348810583895, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.975371360778809, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8672213554382324, "num_tokens": 245242455.0, "step": 6429 }, { "epoch": 0.8179620913369801, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.864154815673828, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.881568193435669, "num_tokens": 245280675.0, "step": 6430 }, { "epoch": 0.8180893016155706, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.950488090515137, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8780961036682129, "num_tokens": 245316947.0, "step": 6431 }, { "epoch": 0.818216511894161, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.93274974822998, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8658401370048523, "num_tokens": 245355145.0, "step": 6432 }, { "epoch": 0.8183437221727515, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.899378776550293, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8736699223518372, "num_tokens": 245390282.0, "step": 6433 }, { "epoch": 0.8184709324513421, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.907212257385254, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8714302182197571, "num_tokens": 245430262.0, "step": 6434 }, { "epoch": 0.8185981427299326, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.941426277160645, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8599870204925537, "num_tokens": 245470874.0, "step": 6435 }, { "epoch": 0.8187253530085231, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.926268577575684, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8680180311203003, "num_tokens": 245506626.0, "step": 6436 }, { "epoch": 0.8188525632871136, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.925928115844727, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8464395999908447, "num_tokens": 245546794.0, "step": 6437 }, { "epoch": 0.8189797735657041, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.921754837036133, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.865647554397583, "num_tokens": 245589544.0, "step": 6438 }, { "epoch": 0.8191069838442946, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.926332473754883, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8784842491149902, "num_tokens": 245628889.0, "step": 6439 }, { "epoch": 0.8192341941228851, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.957297325134277, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8587623834609985, "num_tokens": 245674615.0, "step": 6440 }, { "epoch": 0.8193614044014756, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.946995735168457, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8621048331260681, "num_tokens": 245713966.0, "step": 6441 }, { "epoch": 0.8194886146800662, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.928098678588867, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8671872019767761, "num_tokens": 245750035.0, "step": 6442 }, { "epoch": 0.8196158249586567, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.949067115783691, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8777550458908081, "num_tokens": 245788506.0, "step": 6443 }, { "epoch": 0.8197430352372471, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.91802978515625, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8664653897285461, "num_tokens": 245826547.0, "step": 6444 }, { "epoch": 0.8198702455158376, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.898747444152832, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8528535962104797, "num_tokens": 245874363.0, "step": 6445 }, { "epoch": 0.8199974557944282, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.958866119384766, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.844391405582428, "num_tokens": 245917813.0, "step": 6446 }, { "epoch": 0.8201246660730187, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.930377960205078, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8588268160820007, "num_tokens": 245952177.0, "step": 6447 }, { "epoch": 0.8202518763516092, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.946122169494629, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8646601438522339, "num_tokens": 245992897.0, "step": 6448 }, { "epoch": 0.8203790866301998, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.932674407958984, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.870625913143158, "num_tokens": 246029657.0, "step": 6449 }, { "epoch": 0.8205062969087902, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.913480758666992, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8527082204818726, "num_tokens": 246072425.0, "step": 6450 }, { "epoch": 0.8206335071873807, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.994476318359375, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8800016641616821, "num_tokens": 246113131.0, "step": 6451 }, { "epoch": 0.8207607174659712, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.959527969360352, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8661744594573975, "num_tokens": 246154548.0, "step": 6452 }, { "epoch": 0.8208879277445618, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.975056648254395, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8592977523803711, "num_tokens": 246189518.0, "step": 6453 }, { "epoch": 0.8210151380231523, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.920495986938477, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8580099940299988, "num_tokens": 246235248.0, "step": 6454 }, { "epoch": 0.8211423483017428, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.932082176208496, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8553712368011475, "num_tokens": 246277016.0, "step": 6455 }, { "epoch": 0.8212695585803332, "ewc_loss": 0.02197265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 10.938163757324219, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8597298860549927, "num_tokens": 246312352.0, "step": 6456 }, { "epoch": 0.8213967688589238, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.912545204162598, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8488613367080688, "num_tokens": 246355079.0, "step": 6457 }, { "epoch": 0.8215239791375143, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.930275917053223, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8500314354896545, "num_tokens": 246391073.0, "step": 6458 }, { "epoch": 0.8216511894161048, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.978321075439453, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8689979910850525, "num_tokens": 246435407.0, "step": 6459 }, { "epoch": 0.8217783996946953, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.008195877075195, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8493233323097229, "num_tokens": 246475303.0, "step": 6460 }, { "epoch": 0.8219056099732859, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.901150703430176, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.860253632068634, "num_tokens": 246515237.0, "step": 6461 }, { "epoch": 0.8220328202518764, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.984838485717773, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8570717573165894, "num_tokens": 246554958.0, "step": 6462 }, { "epoch": 0.8221600305304668, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.956098556518555, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8631737232208252, "num_tokens": 246592327.0, "step": 6463 }, { "epoch": 0.8222872408090574, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.949749946594238, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8663709163665771, "num_tokens": 246626965.0, "step": 6464 }, { "epoch": 0.8224144510876479, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.976040840148926, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8594392538070679, "num_tokens": 246662049.0, "step": 6465 }, { "epoch": 0.8225416613662384, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.969611167907715, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8638560175895691, "num_tokens": 246696989.0, "step": 6466 }, { "epoch": 0.8226688716448289, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.001014709472656, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.866657018661499, "num_tokens": 246738418.0, "step": 6467 }, { "epoch": 0.8227960819234195, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.971436500549316, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8636585474014282, "num_tokens": 246776779.0, "step": 6468 }, { "epoch": 0.8229232922020099, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.982168197631836, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8685864210128784, "num_tokens": 246814029.0, "step": 6469 }, { "epoch": 0.8230505024806004, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.017938613891602, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8536428213119507, "num_tokens": 246856390.0, "step": 6470 }, { "epoch": 0.8231777127591909, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.932164192199707, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.863690972328186, "num_tokens": 246895563.0, "step": 6471 }, { "epoch": 0.8233049230377815, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.99984073638916, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8591709136962891, "num_tokens": 246935956.0, "step": 6472 }, { "epoch": 0.823432133316372, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.911172866821289, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8590102791786194, "num_tokens": 246984765.0, "step": 6473 }, { "epoch": 0.8235593435949625, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.011685371398926, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8659951686859131, "num_tokens": 247017086.0, "step": 6474 }, { "epoch": 0.8236865538735529, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.923506736755371, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8678535223007202, "num_tokens": 247056195.0, "step": 6475 }, { "epoch": 0.8238137641521435, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.938887596130371, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8760510683059692, "num_tokens": 247094063.0, "step": 6476 }, { "epoch": 0.823940974430734, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.987053871154785, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8616441488265991, "num_tokens": 247131743.0, "step": 6477 }, { "epoch": 0.8240681847093245, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.9540376663208, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8603826761245728, "num_tokens": 247169477.0, "step": 6478 }, { "epoch": 0.824195394987915, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.937548637390137, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8716601729393005, "num_tokens": 247209159.0, "step": 6479 }, { "epoch": 0.8243226052665056, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.01124095916748, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8636766672134399, "num_tokens": 247247603.0, "step": 6480 }, { "epoch": 0.824449815545096, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.926782608032227, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8728014826774597, "num_tokens": 247290184.0, "step": 6481 }, { "epoch": 0.8245770258236865, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.94566535949707, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8446969985961914, "num_tokens": 247331862.0, "step": 6482 }, { "epoch": 0.824704236102277, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.991124153137207, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8626847863197327, "num_tokens": 247368271.0, "step": 6483 }, { "epoch": 0.8248314463808676, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.920547485351562, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.861634373664856, "num_tokens": 247408077.0, "step": 6484 }, { "epoch": 0.8249586566594581, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.963915824890137, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8697972297668457, "num_tokens": 247447408.0, "step": 6485 }, { "epoch": 0.8250858669380486, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.907023429870605, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8733574151992798, "num_tokens": 247489863.0, "step": 6486 }, { "epoch": 0.8252130772166391, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.963054656982422, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.858725905418396, "num_tokens": 247532233.0, "step": 6487 }, { "epoch": 0.8253402874952296, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.938702583312988, "learning_rate": 1e-06, "loss": 0.5084, "mean_token_accuracy": 0.8391420841217041, "num_tokens": 247570719.0, "step": 6488 }, { "epoch": 0.8254674977738201, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.96461296081543, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8679842948913574, "num_tokens": 247603302.0, "step": 6489 }, { "epoch": 0.8255947080524106, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.9258394241333, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8647110462188721, "num_tokens": 247642837.0, "step": 6490 }, { "epoch": 0.8257219183310012, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.909805297851562, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.859088659286499, "num_tokens": 247685246.0, "step": 6491 }, { "epoch": 0.8258491286095917, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.990535736083984, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8533798456192017, "num_tokens": 247720871.0, "step": 6492 }, { "epoch": 0.8259763388881821, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.914491653442383, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8638272285461426, "num_tokens": 247759796.0, "step": 6493 }, { "epoch": 0.8261035491667726, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.941932678222656, "learning_rate": 1e-06, "loss": 0.5012, "mean_token_accuracy": 0.8420143127441406, "num_tokens": 247806486.0, "step": 6494 }, { "epoch": 0.8262307594453632, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.979724884033203, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8710809946060181, "num_tokens": 247842031.0, "step": 6495 }, { "epoch": 0.8263579697239537, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.94129467010498, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8661448955535889, "num_tokens": 247883438.0, "step": 6496 }, { "epoch": 0.8264851800025442, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.978025436401367, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8705888986587524, "num_tokens": 247921491.0, "step": 6497 }, { "epoch": 0.8266123902811348, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.983481407165527, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8726688027381897, "num_tokens": 247957454.0, "step": 6498 }, { "epoch": 0.8267396005597252, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.953007698059082, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8475207686424255, "num_tokens": 247997705.0, "step": 6499 }, { "epoch": 0.8268668108383157, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.948750495910645, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8575075268745422, "num_tokens": 248034536.0, "step": 6500 }, { "epoch": 0.8269940211169062, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.940468788146973, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8641323447227478, "num_tokens": 248080916.0, "step": 6501 }, { "epoch": 0.8271212313954968, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.996284484863281, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8612267374992371, "num_tokens": 248119719.0, "step": 6502 }, { "epoch": 0.8272484416740873, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.950428009033203, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8592352867126465, "num_tokens": 248156221.0, "step": 6503 }, { "epoch": 0.8273756519526778, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.930052757263184, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8676772117614746, "num_tokens": 248191780.0, "step": 6504 }, { "epoch": 0.8275028622312682, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.983116149902344, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.856762707233429, "num_tokens": 248228594.0, "step": 6505 }, { "epoch": 0.8276300725098588, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.959571838378906, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8668182492256165, "num_tokens": 248260372.0, "step": 6506 }, { "epoch": 0.8277572827884493, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.951129913330078, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8544000387191772, "num_tokens": 248299361.0, "step": 6507 }, { "epoch": 0.8278844930670398, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.987098693847656, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.854960024356842, "num_tokens": 248337370.0, "step": 6508 }, { "epoch": 0.8280117033456303, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.900287628173828, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8645288944244385, "num_tokens": 248377541.0, "step": 6509 }, { "epoch": 0.8281389136242209, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.983583450317383, "learning_rate": 1e-06, "loss": 0.4931, "mean_token_accuracy": 0.845279335975647, "num_tokens": 248413928.0, "step": 6510 }, { "epoch": 0.8282661239028114, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.975384712219238, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8619688153266907, "num_tokens": 248453283.0, "step": 6511 }, { "epoch": 0.8283933341814018, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.936153411865234, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8673352003097534, "num_tokens": 248490286.0, "step": 6512 }, { "epoch": 0.8285205444599923, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.97851848602295, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8473376035690308, "num_tokens": 248526349.0, "step": 6513 }, { "epoch": 0.8286477547385829, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.973669052124023, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8551124334335327, "num_tokens": 248561751.0, "step": 6514 }, { "epoch": 0.8287749650171734, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.904047012329102, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8762100338935852, "num_tokens": 248601921.0, "step": 6515 }, { "epoch": 0.8289021752957639, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.992756843566895, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8731607794761658, "num_tokens": 248641218.0, "step": 6516 }, { "epoch": 0.8290293855743545, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.98932933807373, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.863634467124939, "num_tokens": 248672616.0, "step": 6517 }, { "epoch": 0.8291565958529449, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.974639892578125, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.866308331489563, "num_tokens": 248708212.0, "step": 6518 }, { "epoch": 0.8292838061315354, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.96845531463623, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8673244714736938, "num_tokens": 248746575.0, "step": 6519 }, { "epoch": 0.8294110164101259, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.989214897155762, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8587350845336914, "num_tokens": 248789423.0, "step": 6520 }, { "epoch": 0.8295382266887165, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.987923622131348, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8650586009025574, "num_tokens": 248824044.0, "step": 6521 }, { "epoch": 0.829665436967307, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.960071563720703, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8592160940170288, "num_tokens": 248860805.0, "step": 6522 }, { "epoch": 0.8297926472458975, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.977639198303223, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8686627745628357, "num_tokens": 248898958.0, "step": 6523 }, { "epoch": 0.8299198575244879, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.944647789001465, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8617085218429565, "num_tokens": 248938847.0, "step": 6524 }, { "epoch": 0.8300470678030785, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.999374389648438, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8601455688476562, "num_tokens": 248977944.0, "step": 6525 }, { "epoch": 0.830174278081669, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.97838020324707, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8555928468704224, "num_tokens": 249014610.0, "step": 6526 }, { "epoch": 0.8303014883602595, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.971616744995117, "learning_rate": 1e-06, "loss": 0.5256, "mean_token_accuracy": 0.8330840468406677, "num_tokens": 249053869.0, "step": 6527 }, { "epoch": 0.83042869863885, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.972134590148926, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8550295829772949, "num_tokens": 249096874.0, "step": 6528 }, { "epoch": 0.8305559089174406, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.981222152709961, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8613775968551636, "num_tokens": 249133141.0, "step": 6529 }, { "epoch": 0.830683119196031, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.996203422546387, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8672193288803101, "num_tokens": 249163813.0, "step": 6530 }, { "epoch": 0.8308103294746215, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.934682846069336, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8547946214675903, "num_tokens": 249202935.0, "step": 6531 }, { "epoch": 0.830937539753212, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.98093318939209, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8693214654922485, "num_tokens": 249241061.0, "step": 6532 }, { "epoch": 0.8310647500318026, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.984853744506836, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8510817885398865, "num_tokens": 249277863.0, "step": 6533 }, { "epoch": 0.8311919603103931, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.00660514831543, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8613545894622803, "num_tokens": 249318047.0, "step": 6534 }, { "epoch": 0.8313191705889836, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.968646049499512, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.857196569442749, "num_tokens": 249362095.0, "step": 6535 }, { "epoch": 0.831446380867574, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.978473663330078, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8761730194091797, "num_tokens": 249403362.0, "step": 6536 }, { "epoch": 0.8315735911461646, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.987588882446289, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8691340088844299, "num_tokens": 249438051.0, "step": 6537 }, { "epoch": 0.8317008014247551, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.984328269958496, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8665899038314819, "num_tokens": 249477185.0, "step": 6538 }, { "epoch": 0.8318280117033456, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.015152931213379, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.855189323425293, "num_tokens": 249516880.0, "step": 6539 }, { "epoch": 0.8319552219819362, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.958121299743652, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8720454573631287, "num_tokens": 249555864.0, "step": 6540 }, { "epoch": 0.8320824322605267, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.957297325134277, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8616237044334412, "num_tokens": 249590566.0, "step": 6541 }, { "epoch": 0.8322096425391171, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.974645614624023, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8746463060379028, "num_tokens": 249631523.0, "step": 6542 }, { "epoch": 0.8323368528177076, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.95084285736084, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.870656430721283, "num_tokens": 249674113.0, "step": 6543 }, { "epoch": 0.8324640630962982, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.95917797088623, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8718700408935547, "num_tokens": 249712934.0, "step": 6544 }, { "epoch": 0.8325912733748887, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.94962215423584, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8670033812522888, "num_tokens": 249750295.0, "step": 6545 }, { "epoch": 0.8327184836534792, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.974879264831543, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8668580055236816, "num_tokens": 249790289.0, "step": 6546 }, { "epoch": 0.8328456939320698, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.04088020324707, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8757833242416382, "num_tokens": 249822873.0, "step": 6547 }, { "epoch": 0.8329729042106602, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.985251426696777, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8627705574035645, "num_tokens": 249860564.0, "step": 6548 }, { "epoch": 0.8331001144892507, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.96293830871582, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8618854880332947, "num_tokens": 249901448.0, "step": 6549 }, { "epoch": 0.8332273247678412, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.985740661621094, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8532045483589172, "num_tokens": 249933524.0, "step": 6550 }, { "epoch": 0.8333545350464318, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.000924110412598, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8585342168807983, "num_tokens": 249971856.0, "step": 6551 }, { "epoch": 0.8334817453250223, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.972774505615234, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8544858694076538, "num_tokens": 250013940.0, "step": 6552 }, { "epoch": 0.8336089556036128, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.039813995361328, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.845703125, "num_tokens": 250044826.0, "step": 6553 }, { "epoch": 0.8337361658822032, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.989286422729492, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8470160961151123, "num_tokens": 250087409.0, "step": 6554 }, { "epoch": 0.8338633761607938, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.941864013671875, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8704149723052979, "num_tokens": 250126236.0, "step": 6555 }, { "epoch": 0.8339905864393843, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.018820762634277, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.870290994644165, "num_tokens": 250163559.0, "step": 6556 }, { "epoch": 0.8341177967179748, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.944718360900879, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8633360862731934, "num_tokens": 250203536.0, "step": 6557 }, { "epoch": 0.8342450069965653, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.013638496398926, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8670483827590942, "num_tokens": 250237455.0, "step": 6558 }, { "epoch": 0.8343722172751559, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.950691223144531, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8715044856071472, "num_tokens": 250274559.0, "step": 6559 }, { "epoch": 0.8344994275537464, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.923206329345703, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8647711277008057, "num_tokens": 250309392.0, "step": 6560 }, { "epoch": 0.8346266378323368, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.956005096435547, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8681217432022095, "num_tokens": 250352010.0, "step": 6561 }, { "epoch": 0.8347538481109273, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.953530311584473, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8515125513076782, "num_tokens": 250385745.0, "step": 6562 }, { "epoch": 0.8348810583895179, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.926142692565918, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8607103228569031, "num_tokens": 250426055.0, "step": 6563 }, { "epoch": 0.8350082686681084, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.965757369995117, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8600584268569946, "num_tokens": 250468795.0, "step": 6564 }, { "epoch": 0.8351354789466989, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.935303688049316, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8609800338745117, "num_tokens": 250509115.0, "step": 6565 }, { "epoch": 0.8352626892252895, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.942867279052734, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.862563967704773, "num_tokens": 250547770.0, "step": 6566 }, { "epoch": 0.8353898995038799, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.965409278869629, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8608784675598145, "num_tokens": 250589737.0, "step": 6567 }, { "epoch": 0.8355171097824704, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.960304260253906, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8714663982391357, "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": 11.011129379272461, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8629432916641235, "num_tokens": 250660526.0, "step": 6569 }, { "epoch": 0.8357715303396515, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.985857009887695, "learning_rate": 1e-06, "loss": 0.4883, "mean_token_accuracy": 0.8462995290756226, "num_tokens": 250698218.0, "step": 6570 }, { "epoch": 0.835898740618242, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.999856948852539, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8556002378463745, "num_tokens": 250743410.0, "step": 6571 }, { "epoch": 0.8360259508968325, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.947480201721191, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8722696304321289, "num_tokens": 250781919.0, "step": 6572 }, { "epoch": 0.8361531611754229, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.042276382446289, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8602457046508789, "num_tokens": 250817051.0, "step": 6573 }, { "epoch": 0.8362803714540135, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.999013900756836, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8600990772247314, "num_tokens": 250857241.0, "step": 6574 }, { "epoch": 0.836407581732604, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.010200500488281, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8640021681785583, "num_tokens": 250903591.0, "step": 6575 }, { "epoch": 0.8365347920111945, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.00152587890625, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8564221858978271, "num_tokens": 250945788.0, "step": 6576 }, { "epoch": 0.836662002289785, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.962940216064453, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8554911017417908, "num_tokens": 250984884.0, "step": 6577 }, { "epoch": 0.8367892125683756, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.965188980102539, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8745050430297852, "num_tokens": 251024347.0, "step": 6578 }, { "epoch": 0.836916422846966, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.959403991699219, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8844527006149292, "num_tokens": 251068527.0, "step": 6579 }, { "epoch": 0.8370436331255565, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.95323657989502, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8680012226104736, "num_tokens": 251109914.0, "step": 6580 }, { "epoch": 0.837170843404147, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.02871322631836, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8549031019210815, "num_tokens": 251144760.0, "step": 6581 }, { "epoch": 0.8372980536827376, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.958124160766602, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8607367277145386, "num_tokens": 251184456.0, "step": 6582 }, { "epoch": 0.8374252639613281, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.965874671936035, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8551977276802063, "num_tokens": 251229493.0, "step": 6583 }, { "epoch": 0.8375524742399186, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.0135498046875, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.867068886756897, "num_tokens": 251270886.0, "step": 6584 }, { "epoch": 0.837679684518509, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.960675239562988, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8712311387062073, "num_tokens": 251304833.0, "step": 6585 }, { "epoch": 0.8378068947970996, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.039336204528809, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8646520376205444, "num_tokens": 251334196.0, "step": 6586 }, { "epoch": 0.8379341050756901, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.998088836669922, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.85255366563797, "num_tokens": 251380278.0, "step": 6587 }, { "epoch": 0.8380613153542806, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.958393096923828, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8577330112457275, "num_tokens": 251418432.0, "step": 6588 }, { "epoch": 0.8381885256328712, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.977713584899902, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8602964282035828, "num_tokens": 251453747.0, "step": 6589 }, { "epoch": 0.8383157359114617, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.963664054870605, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8630335330963135, "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": 11.01626205444336, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8636142611503601, "num_tokens": 251527945.0, "step": 6591 }, { "epoch": 0.8385701564686426, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.960517883300781, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8709259629249573, "num_tokens": 251567161.0, "step": 6592 }, { "epoch": 0.8386973667472332, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.997110366821289, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.861634373664856, "num_tokens": 251601189.0, "step": 6593 }, { "epoch": 0.8388245770258237, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.061810493469238, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8600857257843018, "num_tokens": 251630846.0, "step": 6594 }, { "epoch": 0.8389517873044142, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.974449157714844, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8701525926589966, "num_tokens": 251665701.0, "step": 6595 }, { "epoch": 0.8390789975830047, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.027116775512695, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8556717038154602, "num_tokens": 251699909.0, "step": 6596 }, { "epoch": 0.8392062078615952, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.031852722167969, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.860077440738678, "num_tokens": 251734052.0, "step": 6597 }, { "epoch": 0.8393334181401857, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.959942817687988, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8514007329940796, "num_tokens": 251772518.0, "step": 6598 }, { "epoch": 0.8394606284187762, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.008322715759277, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8605126142501831, "num_tokens": 251816378.0, "step": 6599 }, { "epoch": 0.8395878386973668, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.986916542053223, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.862446665763855, "num_tokens": 251851264.0, "step": 6600 }, { "epoch": 0.8397150489759573, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.998624801635742, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8647321462631226, "num_tokens": 251888906.0, "step": 6601 }, { "epoch": 0.8398422592545478, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.96607494354248, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8694317936897278, "num_tokens": 251930641.0, "step": 6602 }, { "epoch": 0.8399694695331382, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.964709281921387, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8438658118247986, "num_tokens": 251968929.0, "step": 6603 }, { "epoch": 0.8400966798117288, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.016733169555664, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8542169332504272, "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": 11.017857551574707, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8584800958633423, "num_tokens": 252045861.0, "step": 6605 }, { "epoch": 0.8403511003689098, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.058676719665527, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8552613854408264, "num_tokens": 252084412.0, "step": 6606 }, { "epoch": 0.8404783106475003, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.990915298461914, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8669921159744263, "num_tokens": 252120576.0, "step": 6607 }, { "epoch": 0.8406055209260909, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.000996589660645, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8601540923118591, "num_tokens": 252161636.0, "step": 6608 }, { "epoch": 0.8407327312046813, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.978229522705078, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8626402616500854, "num_tokens": 252199669.0, "step": 6609 }, { "epoch": 0.8408599414832718, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.995635032653809, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8692335486412048, "num_tokens": 252236832.0, "step": 6610 }, { "epoch": 0.8409871517618623, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.984850883483887, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8580924868583679, "num_tokens": 252282518.0, "step": 6611 }, { "epoch": 0.8411143620404529, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.065985679626465, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8560800552368164, "num_tokens": 252316479.0, "step": 6612 }, { "epoch": 0.8412415723190434, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.97588062286377, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8562017679214478, "num_tokens": 252353789.0, "step": 6613 }, { "epoch": 0.8413687825976339, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.974520683288574, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8657068014144897, "num_tokens": 252391609.0, "step": 6614 }, { "epoch": 0.8414959928762245, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.953725814819336, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8759787082672119, "num_tokens": 252433514.0, "step": 6615 }, { "epoch": 0.8416232031548149, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.976371765136719, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8659329414367676, "num_tokens": 252469700.0, "step": 6616 }, { "epoch": 0.8417504134334054, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.027559280395508, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8582948446273804, "num_tokens": 252501658.0, "step": 6617 }, { "epoch": 0.8418776237119959, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.061090469360352, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8645621538162231, "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": 10.949589729309082, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8701278567314148, "num_tokens": 252573172.0, "step": 6619 }, { "epoch": 0.842132044269177, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.003652572631836, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8715137243270874, "num_tokens": 252607258.0, "step": 6620 }, { "epoch": 0.8422592545477675, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.987109184265137, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8733546733856201, "num_tokens": 252651451.0, "step": 6621 }, { "epoch": 0.8423864648263579, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.963871002197266, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8748376965522766, "num_tokens": 252686348.0, "step": 6622 }, { "epoch": 0.8425136751049485, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.035109519958496, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8575206995010376, "num_tokens": 252724316.0, "step": 6623 }, { "epoch": 0.842640885383539, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.995912551879883, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8516607880592346, "num_tokens": 252764612.0, "step": 6624 }, { "epoch": 0.8427680956621295, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.953113555908203, "learning_rate": 1e-06, "loss": 0.4898, "mean_token_accuracy": 0.8420522809028625, "num_tokens": 252809510.0, "step": 6625 }, { "epoch": 0.84289530594072, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.05101490020752, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8599334955215454, "num_tokens": 252847598.0, "step": 6626 }, { "epoch": 0.8430225162193106, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.031712532043457, "learning_rate": 1e-06, "loss": 0.4899, "mean_token_accuracy": 0.8482074737548828, "num_tokens": 252884401.0, "step": 6627 }, { "epoch": 0.843149726497901, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.002814292907715, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8468611240386963, "num_tokens": 252928348.0, "step": 6628 }, { "epoch": 0.8432769367764915, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.057578086853027, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8622278571128845, "num_tokens": 252965084.0, "step": 6629 }, { "epoch": 0.843404147055082, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.963764190673828, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8670890927314758, "num_tokens": 253003697.0, "step": 6630 }, { "epoch": 0.8435313573336726, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.041471481323242, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8485654592514038, "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": 10.943035125732422, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8536971807479858, "num_tokens": 253078163.0, "step": 6632 }, { "epoch": 0.8437857778908536, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.999618530273438, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.860578715801239, "num_tokens": 253118907.0, "step": 6633 }, { "epoch": 0.843912988169444, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.058382987976074, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8809623718261719, "num_tokens": 253153132.0, "step": 6634 }, { "epoch": 0.8440401984480346, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.044368743896484, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8584709763526917, "num_tokens": 253194446.0, "step": 6635 }, { "epoch": 0.8441674087266251, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.0139741897583, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.864248514175415, "num_tokens": 253230798.0, "step": 6636 }, { "epoch": 0.8442946190052156, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.006399154663086, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8547372817993164, "num_tokens": 253269029.0, "step": 6637 }, { "epoch": 0.8444218292838062, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.008970260620117, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8575039505958557, "num_tokens": 253310656.0, "step": 6638 }, { "epoch": 0.8445490395623967, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.019335746765137, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.880229115486145, "num_tokens": 253346882.0, "step": 6639 }, { "epoch": 0.8446762498409871, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.09708309173584, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8523260951042175, "num_tokens": 253389601.0, "step": 6640 }, { "epoch": 0.8448034601195776, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.947879791259766, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.860708475112915, "num_tokens": 253428725.0, "step": 6641 }, { "epoch": 0.8449306703981682, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.041579246520996, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8768163323402405, "num_tokens": 253461404.0, "step": 6642 }, { "epoch": 0.8450578806767587, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.0199556350708, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8667081594467163, "num_tokens": 253500324.0, "step": 6643 }, { "epoch": 0.8451850909553492, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.98665714263916, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8501803874969482, "num_tokens": 253542500.0, "step": 6644 }, { "epoch": 0.8453123012339397, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.030069351196289, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8447161912918091, "num_tokens": 253581887.0, "step": 6645 }, { "epoch": 0.8454395115125302, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.037897109985352, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8708033561706543, "num_tokens": 253616942.0, "step": 6646 }, { "epoch": 0.8455667217911207, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.010629653930664, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8560656309127808, "num_tokens": 253652610.0, "step": 6647 }, { "epoch": 0.8456939320697112, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.101832389831543, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8599156141281128, "num_tokens": 253686269.0, "step": 6648 }, { "epoch": 0.8458211423483017, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.972555160522461, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8571017980575562, "num_tokens": 253723632.0, "step": 6649 }, { "epoch": 0.8459483526268923, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.080947875976562, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.855384111404419, "num_tokens": 253755990.0, "step": 6650 }, { "epoch": 0.8460755629054828, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.032021522521973, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8564014434814453, "num_tokens": 253794200.0, "step": 6651 }, { "epoch": 0.8462027731840732, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.996767044067383, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.873058557510376, "num_tokens": 253840554.0, "step": 6652 }, { "epoch": 0.8463299834626637, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.044715881347656, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8722073435783386, "num_tokens": 253873642.0, "step": 6653 }, { "epoch": 0.8464571937412543, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.043035507202148, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8672114014625549, "num_tokens": 253911839.0, "step": 6654 }, { "epoch": 0.8465844040198448, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.04690933227539, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.852060079574585, "num_tokens": 253944649.0, "step": 6655 }, { "epoch": 0.8467116142984353, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 27.98735237121582, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8620837926864624, "num_tokens": 253977739.0, "step": 6656 }, { "epoch": 0.8468388245770259, "ewc_loss": 0.03271484375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.208045959472656, "learning_rate": 1e-06, "loss": 0.4767, "mean_token_accuracy": 0.8474081754684448, "num_tokens": 254021216.0, "step": 6657 }, { "epoch": 0.8469660348556163, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 9.635584831237793, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8596372008323669, "num_tokens": 254063209.0, "step": 6658 }, { "epoch": 0.8470932451342068, "ewc_loss": 0.02490234375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.638124465942383, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.857776403427124, "num_tokens": 254102325.0, "step": 6659 }, { "epoch": 0.8472204554127973, "ewc_loss": 0.02880859375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 12.435980796813965, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8671663999557495, "num_tokens": 254136991.0, "step": 6660 }, { "epoch": 0.8473476656913879, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.292742729187012, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8616428375244141, "num_tokens": 254179520.0, "step": 6661 }, { "epoch": 0.8474748759699784, "ewc_loss": 0.025634765625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 11.743672370910645, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8630545735359192, "num_tokens": 254220661.0, "step": 6662 }, { "epoch": 0.8476020862485689, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 11.573076248168945, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8694385290145874, "num_tokens": 254261853.0, "step": 6663 }, { "epoch": 0.8477292965271594, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.526686668395996, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8623883724212646, "num_tokens": 254301577.0, "step": 6664 }, { "epoch": 0.8478565068057499, "ewc_loss": 0.025390625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.431184768676758, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8748251795768738, "num_tokens": 254341792.0, "step": 6665 }, { "epoch": 0.8479837170843404, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.422605514526367, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.849371075630188, "num_tokens": 254376755.0, "step": 6666 }, { "epoch": 0.8481109273629309, "ewc_loss": 0.024658203125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.300192832946777, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8735999464988708, "num_tokens": 254413565.0, "step": 6667 }, { "epoch": 0.8482381376415215, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.24254322052002, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.850829541683197, "num_tokens": 254449889.0, "step": 6668 }, { "epoch": 0.848365347920112, "ewc_loss": 0.024169921875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.223201751708984, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8758403658866882, "num_tokens": 254488129.0, "step": 6669 }, { "epoch": 0.8484925581987025, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.180204391479492, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8725688457489014, "num_tokens": 254522031.0, "step": 6670 }, { "epoch": 0.8486197684772929, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.118535041809082, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8740999102592468, "num_tokens": 254559249.0, "step": 6671 }, { "epoch": 0.8487469787558835, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.08437442779541, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8555247783660889, "num_tokens": 254603968.0, "step": 6672 }, { "epoch": 0.848874189034474, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.07042407989502, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8716297745704651, "num_tokens": 254642178.0, "step": 6673 }, { "epoch": 0.8490013993130645, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.079413414001465, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8541792631149292, "num_tokens": 254679129.0, "step": 6674 }, { "epoch": 0.849128609591655, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.151391983032227, "learning_rate": 1e-06, "loss": 0.5322, "mean_token_accuracy": 0.8331903219223022, "num_tokens": 254714952.0, "step": 6675 }, { "epoch": 0.8492558198702456, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.062524795532227, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8578442335128784, "num_tokens": 254743323.0, "step": 6676 }, { "epoch": 0.849383030148836, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.051094055175781, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8690392971038818, "num_tokens": 254784749.0, "step": 6677 }, { "epoch": 0.8495102404274265, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 10.98908805847168, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8684003353118896, "num_tokens": 254820102.0, "step": 6678 }, { "epoch": 0.849637450706017, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.024819374084473, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8579933047294617, "num_tokens": 254859988.0, "step": 6679 }, { "epoch": 0.8497646609846076, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 10.939297676086426, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8575860261917114, "num_tokens": 254906792.0, "step": 6680 }, { "epoch": 0.8498918712631981, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.014076232910156, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8646652698516846, "num_tokens": 254941552.0, "step": 6681 }, { "epoch": 0.8500190815417886, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.012452125549316, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8551824688911438, "num_tokens": 254974653.0, "step": 6682 }, { "epoch": 0.850146291820379, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.972804069519043, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8470138311386108, "num_tokens": 255018418.0, "step": 6683 }, { "epoch": 0.8502735020989696, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.962467193603516, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8677136898040771, "num_tokens": 255049626.0, "step": 6684 }, { "epoch": 0.8504007123775601, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.990612983703613, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8555343151092529, "num_tokens": 255086323.0, "step": 6685 }, { "epoch": 0.8505279226561506, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.00615119934082, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8697996139526367, "num_tokens": 255124102.0, "step": 6686 }, { "epoch": 0.8506551329347412, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.000446319580078, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8660362958908081, "num_tokens": 255162176.0, "step": 6687 }, { "epoch": 0.8507823432133317, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.051300048828125, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.865812361240387, "num_tokens": 255196256.0, "step": 6688 }, { "epoch": 0.8509095534919221, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.950782775878906, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8548704385757446, "num_tokens": 255236752.0, "step": 6689 }, { "epoch": 0.8510367637705126, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.022969245910645, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8630681037902832, "num_tokens": 255268694.0, "step": 6690 }, { "epoch": 0.8511639740491032, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.98846435546875, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8682315349578857, "num_tokens": 255309445.0, "step": 6691 }, { "epoch": 0.8512911843276937, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.996664047241211, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8654050827026367, "num_tokens": 255341126.0, "step": 6692 }, { "epoch": 0.8514183946062842, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.941802978515625, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8703177571296692, "num_tokens": 255379553.0, "step": 6693 }, { "epoch": 0.8515456048848747, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.979150772094727, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8628356456756592, "num_tokens": 255416588.0, "step": 6694 }, { "epoch": 0.8516728151634652, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.992555618286133, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8503417372703552, "num_tokens": 255449574.0, "step": 6695 }, { "epoch": 0.8518000254420557, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.920166969299316, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8602319359779358, "num_tokens": 255494088.0, "step": 6696 }, { "epoch": 0.8519272357206462, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.00393295288086, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.871130108833313, "num_tokens": 255529785.0, "step": 6697 }, { "epoch": 0.8520544459992367, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.986916542053223, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8492639660835266, "num_tokens": 255566675.0, "step": 6698 }, { "epoch": 0.8521816562778273, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00058364868164, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8664630651473999, "num_tokens": 255606502.0, "step": 6699 }, { "epoch": 0.8523088665564178, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.974600791931152, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8533554077148438, "num_tokens": 255649179.0, "step": 6700 }, { "epoch": 0.8524360768350082, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.014030456542969, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8612691164016724, "num_tokens": 255688190.0, "step": 6701 }, { "epoch": 0.8525632871135987, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.985896110534668, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8611756563186646, "num_tokens": 255730401.0, "step": 6702 }, { "epoch": 0.8526904973921893, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.983262062072754, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8581308126449585, "num_tokens": 255767881.0, "step": 6703 }, { "epoch": 0.8528177076707798, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.985383033752441, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8554290533065796, "num_tokens": 255808347.0, "step": 6704 }, { "epoch": 0.8529449179493703, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.936036109924316, "learning_rate": 1e-06, "loss": 0.4922, "mean_token_accuracy": 0.8436747789382935, "num_tokens": 255849538.0, "step": 6705 }, { "epoch": 0.8530721282279609, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.987323760986328, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8720738291740417, "num_tokens": 255890228.0, "step": 6706 }, { "epoch": 0.8531993385065513, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.957200050354004, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8609632849693298, "num_tokens": 255930189.0, "step": 6707 }, { "epoch": 0.8533265487851418, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.985992431640625, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8684139847755432, "num_tokens": 255968310.0, "step": 6708 }, { "epoch": 0.8534537590637323, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.96779727935791, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8732678294181824, "num_tokens": 256005872.0, "step": 6709 }, { "epoch": 0.8535809693423229, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.988301277160645, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8677496910095215, "num_tokens": 256045374.0, "step": 6710 }, { "epoch": 0.8537081796209134, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.97508430480957, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8655728697776794, "num_tokens": 256084874.0, "step": 6711 }, { "epoch": 0.8538353898995039, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.998137474060059, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8552172780036926, "num_tokens": 256122251.0, "step": 6712 }, { "epoch": 0.8539626001780944, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.002890586853027, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8600687384605408, "num_tokens": 256165025.0, "step": 6713 }, { "epoch": 0.8540898104566849, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.018381118774414, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8688867688179016, "num_tokens": 256202220.0, "step": 6714 }, { "epoch": 0.8542170207352754, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.01655387878418, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.865565299987793, "num_tokens": 256234036.0, "step": 6715 }, { "epoch": 0.8543442310138659, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.011100769042969, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8604016900062561, "num_tokens": 256275253.0, "step": 6716 }, { "epoch": 0.8544714412924564, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.976408004760742, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8599525094032288, "num_tokens": 256311296.0, "step": 6717 }, { "epoch": 0.854598651571047, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.995600700378418, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8688005208969116, "num_tokens": 256348149.0, "step": 6718 }, { "epoch": 0.8547258618496375, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00717544555664, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8760254979133606, "num_tokens": 256387861.0, "step": 6719 }, { "epoch": 0.8548530721282279, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.01176643371582, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8706731200218201, "num_tokens": 256427131.0, "step": 6720 }, { "epoch": 0.8549802824068184, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.987676620483398, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8673166632652283, "num_tokens": 256465041.0, "step": 6721 }, { "epoch": 0.855107492685409, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.969810485839844, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8535287976264954, "num_tokens": 256502576.0, "step": 6722 }, { "epoch": 0.8552347029639995, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.026397705078125, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8646194338798523, "num_tokens": 256540493.0, "step": 6723 }, { "epoch": 0.85536191324259, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.03046703338623, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8450376987457275, "num_tokens": 256578115.0, "step": 6724 }, { "epoch": 0.8554891235211806, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.992439270019531, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8692831993103027, "num_tokens": 256615957.0, "step": 6725 }, { "epoch": 0.855616333799771, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.000691413879395, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8660145998001099, "num_tokens": 256649972.0, "step": 6726 }, { "epoch": 0.8557435440783615, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 10.996635437011719, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8699257373809814, "num_tokens": 256686683.0, "step": 6727 }, { "epoch": 0.855870754356952, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.045522689819336, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8607304692268372, "num_tokens": 256723587.0, "step": 6728 }, { "epoch": 0.8559979646355426, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00034236907959, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8633014559745789, "num_tokens": 256765245.0, "step": 6729 }, { "epoch": 0.8561251749141331, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.953387260437012, "learning_rate": 1e-06, "loss": 0.3335, "mean_token_accuracy": 0.8903182148933411, "num_tokens": 256806930.0, "step": 6730 }, { "epoch": 0.8562523851927236, "ewc_loss": 0.022216796875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.028717994689941, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8604666590690613, "num_tokens": 256844568.0, "step": 6731 }, { "epoch": 0.856379595471314, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.06279468536377, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8610488176345825, "num_tokens": 256877483.0, "step": 6732 }, { "epoch": 0.8565068057499046, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.012471199035645, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8674224615097046, "num_tokens": 256921139.0, "step": 6733 }, { "epoch": 0.8566340160284951, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.003656387329102, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.862321138381958, "num_tokens": 256957861.0, "step": 6734 }, { "epoch": 0.8567612263070856, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.017024040222168, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8644169569015503, "num_tokens": 256998582.0, "step": 6735 }, { "epoch": 0.8568884365856761, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.985331535339355, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8616576790809631, "num_tokens": 257037638.0, "step": 6736 }, { "epoch": 0.8570156468642667, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.935103416442871, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8649939894676208, "num_tokens": 257074427.0, "step": 6737 }, { "epoch": 0.8571428571428571, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.024857521057129, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8655209541320801, "num_tokens": 257113826.0, "step": 6738 }, { "epoch": 0.8572700674214476, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.98023796081543, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8492584228515625, "num_tokens": 257152387.0, "step": 6739 }, { "epoch": 0.8573972777000382, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.0035982131958, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.864156186580658, "num_tokens": 257192083.0, "step": 6740 }, { "epoch": 0.8575244879786287, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.950543403625488, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8551108837127686, "num_tokens": 257235425.0, "step": 6741 }, { "epoch": 0.8576516982572192, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.014850616455078, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8586490154266357, "num_tokens": 257268621.0, "step": 6742 }, { "epoch": 0.8577789085358097, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.963414192199707, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8677270412445068, "num_tokens": 257306909.0, "step": 6743 }, { "epoch": 0.8579061188144002, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.985926628112793, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.877088725566864, "num_tokens": 257343388.0, "step": 6744 }, { "epoch": 0.8580333290929907, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.999874114990234, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8681937456130981, "num_tokens": 257377172.0, "step": 6745 }, { "epoch": 0.8581605393715812, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.993902206420898, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8562034368515015, "num_tokens": 257410312.0, "step": 6746 }, { "epoch": 0.8582877496501717, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.957171440124512, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.866222620010376, "num_tokens": 257449919.0, "step": 6747 }, { "epoch": 0.8584149599287623, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.99923324584961, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8741264343261719, "num_tokens": 257488296.0, "step": 6748 }, { "epoch": 0.8585421702073528, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.981057167053223, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8720695972442627, "num_tokens": 257524678.0, "step": 6749 }, { "epoch": 0.8586693804859432, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.951074600219727, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8682996034622192, "num_tokens": 257559060.0, "step": 6750 }, { "epoch": 0.8587965907645337, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.98294734954834, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8725890517234802, "num_tokens": 257594340.0, "step": 6751 }, { "epoch": 0.8589238010431243, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.959138870239258, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8745409250259399, "num_tokens": 257627936.0, "step": 6752 }, { "epoch": 0.8590510113217148, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.016722679138184, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8606320023536682, "num_tokens": 257660311.0, "step": 6753 }, { "epoch": 0.8591782216003053, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.01932144165039, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8555614948272705, "num_tokens": 257697496.0, "step": 6754 }, { "epoch": 0.8593054318788959, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.005995750427246, "learning_rate": 1e-06, "loss": 0.4804, "mean_token_accuracy": 0.8455228805541992, "num_tokens": 257737242.0, "step": 6755 }, { "epoch": 0.8594326421574863, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.027287483215332, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8639692068099976, "num_tokens": 257772946.0, "step": 6756 }, { "epoch": 0.8595598524360768, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.03249454498291, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.857041597366333, "num_tokens": 257807198.0, "step": 6757 }, { "epoch": 0.8596870627146673, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.010190963745117, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8697551488876343, "num_tokens": 257848618.0, "step": 6758 }, { "epoch": 0.8598142729932579, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.011194229125977, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8751709461212158, "num_tokens": 257887805.0, "step": 6759 }, { "epoch": 0.8599414832718484, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.973169326782227, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8689131140708923, "num_tokens": 257929276.0, "step": 6760 }, { "epoch": 0.8600686935504389, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.027698516845703, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8576083779335022, "num_tokens": 257958629.0, "step": 6761 }, { "epoch": 0.8601959038290294, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.077342987060547, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.859568178653717, "num_tokens": 257990303.0, "step": 6762 }, { "epoch": 0.8603231141076199, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.983344078063965, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8464725613594055, "num_tokens": 258028704.0, "step": 6763 }, { "epoch": 0.8604503243862104, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.025656700134277, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8714715242385864, "num_tokens": 258062757.0, "step": 6764 }, { "epoch": 0.8605775346648009, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.052964210510254, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8626053333282471, "num_tokens": 258098574.0, "step": 6765 }, { "epoch": 0.8607047449433914, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.038000106811523, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8720475435256958, "num_tokens": 258130937.0, "step": 6766 }, { "epoch": 0.860831955221982, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.067086219787598, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8661219477653503, "num_tokens": 258165142.0, "step": 6767 }, { "epoch": 0.8609591655005725, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.94793701171875, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8803467750549316, "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": 11.069367408752441, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.864195704460144, "num_tokens": 258238692.0, "step": 6769 }, { "epoch": 0.8612135860577534, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.068999290466309, "learning_rate": 1e-06, "loss": 0.4963, "mean_token_accuracy": 0.8440116047859192, "num_tokens": 258280942.0, "step": 6770 }, { "epoch": 0.861340796336344, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.057812690734863, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8663791418075562, "num_tokens": 258318835.0, "step": 6771 }, { "epoch": 0.8614680066149345, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.088138580322266, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8545913696289062, "num_tokens": 258359461.0, "step": 6772 }, { "epoch": 0.861595216893525, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.016858100891113, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.859211802482605, "num_tokens": 258394582.0, "step": 6773 }, { "epoch": 0.8617224271721156, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.05016040802002, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8662599921226501, "num_tokens": 258438526.0, "step": 6774 }, { "epoch": 0.861849637450706, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.999405860900879, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8505549430847168, "num_tokens": 258479332.0, "step": 6775 }, { "epoch": 0.8619768477292965, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.081063270568848, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8534743785858154, "num_tokens": 258513202.0, "step": 6776 }, { "epoch": 0.862104058007887, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.991344451904297, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8650517463684082, "num_tokens": 258549145.0, "step": 6777 }, { "epoch": 0.8622312682864776, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.004426956176758, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8545719981193542, "num_tokens": 258586421.0, "step": 6778 }, { "epoch": 0.8623584785650681, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.04526424407959, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8638139963150024, "num_tokens": 258620014.0, "step": 6779 }, { "epoch": 0.8624856888436586, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.017898559570312, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8689914345741272, "num_tokens": 258656124.0, "step": 6780 }, { "epoch": 0.862612899122249, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.01257610321045, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8525044918060303, "num_tokens": 258692537.0, "step": 6781 }, { "epoch": 0.8627401094008396, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.006490707397461, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8663987517356873, "num_tokens": 258732737.0, "step": 6782 }, { "epoch": 0.8628673196794301, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.999714851379395, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8506287336349487, "num_tokens": 258770634.0, "step": 6783 }, { "epoch": 0.8629945299580206, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.026116371154785, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8623807430267334, "num_tokens": 258809643.0, "step": 6784 }, { "epoch": 0.8631217402366111, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.04944896697998, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8682947158813477, "num_tokens": 258845336.0, "step": 6785 }, { "epoch": 0.8632489505152017, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.023184776306152, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8645581007003784, "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": 11.006326675415039, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.870570182800293, "num_tokens": 258918661.0, "step": 6787 }, { "epoch": 0.8635033710723826, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.00200080871582, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8582935333251953, "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": 10.970945358276367, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8543291091918945, "num_tokens": 259003507.0, "step": 6789 }, { "epoch": 0.8637577916295637, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.053621292114258, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.857897162437439, "num_tokens": 259042284.0, "step": 6790 }, { "epoch": 0.8638850019081542, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.96018123626709, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8767520189285278, "num_tokens": 259082929.0, "step": 6791 }, { "epoch": 0.8640122121867447, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.018362998962402, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8783518075942993, "num_tokens": 259125920.0, "step": 6792 }, { "epoch": 0.8641394224653351, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.015768051147461, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8644814491271973, "num_tokens": 259163140.0, "step": 6793 }, { "epoch": 0.8642666327439257, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.024888038635254, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.861179769039154, "num_tokens": 259199038.0, "step": 6794 }, { "epoch": 0.8643938430225162, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.0700044631958, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8614770174026489, "num_tokens": 259235713.0, "step": 6795 }, { "epoch": 0.8645210533011067, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.010359764099121, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8527795076370239, "num_tokens": 259278314.0, "step": 6796 }, { "epoch": 0.8646482635796973, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.005297660827637, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8718427419662476, "num_tokens": 259313039.0, "step": 6797 }, { "epoch": 0.8647754738582878, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.101102828979492, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.869494616985321, "num_tokens": 259348845.0, "step": 6798 }, { "epoch": 0.8649026841368782, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.011414527893066, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8575010895729065, "num_tokens": 259387345.0, "step": 6799 }, { "epoch": 0.8650298944154687, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.025983810424805, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8625096082687378, "num_tokens": 259427835.0, "step": 6800 }, { "epoch": 0.8651571046940593, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.041921615600586, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8670126795768738, "num_tokens": 259460453.0, "step": 6801 }, { "epoch": 0.8652843149726498, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.993270874023438, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8698054552078247, "num_tokens": 259499296.0, "step": 6802 }, { "epoch": 0.8654115252512403, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.99191665649414, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8602632284164429, "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": 11.009387016296387, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8556749224662781, "num_tokens": 259583136.0, "step": 6804 }, { "epoch": 0.8656659458084213, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.999670028686523, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.867316484451294, "num_tokens": 259620422.0, "step": 6805 }, { "epoch": 0.8657931560870118, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.082250595092773, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8504973649978638, "num_tokens": 259663948.0, "step": 6806 }, { "epoch": 0.8659203663656023, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.931072235107422, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.864827573299408, "num_tokens": 259701986.0, "step": 6807 }, { "epoch": 0.8660475766441929, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.08617877960205, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8782145977020264, "num_tokens": 259740684.0, "step": 6808 }, { "epoch": 0.8661747869227834, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.994924545288086, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8465416431427002, "num_tokens": 259781659.0, "step": 6809 }, { "epoch": 0.8663019972013739, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 10.995915412902832, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8570092916488647, "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": 11.023187637329102, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8647318482398987, "num_tokens": 259856652.0, "step": 6811 }, { "epoch": 0.8665564177585549, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.994691848754883, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8811835646629333, "num_tokens": 259898255.0, "step": 6812 }, { "epoch": 0.8666836280371454, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 11.030298233032227, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8679078221321106, "num_tokens": 259935305.0, "step": 6813 }, { "epoch": 0.8668108383157359, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.035655975341797, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8676317930221558, "num_tokens": 259975245.0, "step": 6814 }, { "epoch": 0.8669380485943264, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.01040267944336, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8687788844108582, "num_tokens": 260014150.0, "step": 6815 }, { "epoch": 0.867065258872917, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.014701843261719, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8538602590560913, "num_tokens": 260054655.0, "step": 6816 }, { "epoch": 0.8671924691515075, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.99084758758545, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8642135858535767, "num_tokens": 260091528.0, "step": 6817 }, { "epoch": 0.8673196794300979, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.016247749328613, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8777554631233215, "num_tokens": 260123069.0, "step": 6818 }, { "epoch": 0.8674468897086884, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.06513786315918, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8484770059585571, "num_tokens": 260154482.0, "step": 6819 }, { "epoch": 0.867574099987279, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.014769554138184, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8654264211654663, "num_tokens": 260193983.0, "step": 6820 }, { "epoch": 0.8677013102658695, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.025339126586914, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8654512763023376, "num_tokens": 260228705.0, "step": 6821 }, { "epoch": 0.86782852054446, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.073203086853027, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8620983362197876, "num_tokens": 260262821.0, "step": 6822 }, { "epoch": 0.8679557308230506, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.98091983795166, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8616523146629333, "num_tokens": 260298949.0, "step": 6823 }, { "epoch": 0.868082941101641, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.021578788757324, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8709326982498169, "num_tokens": 260340028.0, "step": 6824 }, { "epoch": 0.8682101513802315, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.983065605163574, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8631880879402161, "num_tokens": 260381328.0, "step": 6825 }, { "epoch": 0.868337361658822, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.009239196777344, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8636212348937988, "num_tokens": 260423421.0, "step": 6826 }, { "epoch": 0.8684645719374126, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.008943557739258, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8617122173309326, "num_tokens": 260461985.0, "step": 6827 }, { "epoch": 0.8685917822160031, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.004888534545898, "learning_rate": 1e-06, "loss": 0.4845, "mean_token_accuracy": 0.8483047485351562, "num_tokens": 260503041.0, "step": 6828 }, { "epoch": 0.8687189924945936, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.997130393981934, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8715068101882935, "num_tokens": 260545553.0, "step": 6829 }, { "epoch": 0.868846202773184, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.0829496383667, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8704384565353394, "num_tokens": 260577160.0, "step": 6830 }, { "epoch": 0.8689734130517746, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 10.967632293701172, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8744895458221436, "num_tokens": 260614748.0, "step": 6831 }, { "epoch": 0.8691006233303651, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.007302284240723, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.863961935043335, "num_tokens": 260651762.0, "step": 6832 }, { "epoch": 0.8692278336089556, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.065202713012695, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8519542813301086, "num_tokens": 260694438.0, "step": 6833 }, { "epoch": 0.8693550438875461, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.987637519836426, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8716326951980591, "num_tokens": 260730256.0, "step": 6834 }, { "epoch": 0.8694822541661367, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.041146278381348, "learning_rate": 1e-06, "loss": 0.4902, "mean_token_accuracy": 0.8428427577018738, "num_tokens": 260766190.0, "step": 6835 }, { "epoch": 0.8696094644447271, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.016332626342773, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.877678394317627, "num_tokens": 260801733.0, "step": 6836 }, { "epoch": 0.8697366747233176, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.030061721801758, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8745825886726379, "num_tokens": 260843461.0, "step": 6837 }, { "epoch": 0.8698638850019081, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.064013481140137, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8615816235542297, "num_tokens": 260875006.0, "step": 6838 }, { "epoch": 0.8699910952804987, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 10.984707832336426, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8575292229652405, "num_tokens": 260917553.0, "step": 6839 }, { "epoch": 0.8701183055590892, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.067748069763184, "learning_rate": 1e-06, "loss": 0.4977, "mean_token_accuracy": 0.8386144042015076, "num_tokens": 260955293.0, "step": 6840 }, { "epoch": 0.8702455158376797, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.096437454223633, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.863311231136322, "num_tokens": 260983966.0, "step": 6841 }, { "epoch": 0.8703727261162701, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.054712295532227, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8765546679496765, "num_tokens": 261019434.0, "step": 6842 }, { "epoch": 0.8704999363948607, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.054699897766113, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8522644639015198, "num_tokens": 261053761.0, "step": 6843 }, { "epoch": 0.8706271466734512, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.10769271850586, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8532464504241943, "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": 11.068767547607422, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8642455339431763, "num_tokens": 261129518.0, "step": 6845 }, { "epoch": 0.8708815672306323, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.090531349182129, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8459531664848328, "num_tokens": 261164282.0, "step": 6846 }, { "epoch": 0.8710087775092228, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.081337928771973, "learning_rate": 1e-06, "loss": 0.4902, "mean_token_accuracy": 0.8434287309646606, "num_tokens": 261205055.0, "step": 6847 }, { "epoch": 0.8711359877878132, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.079852104187012, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8581721186637878, "num_tokens": 261242329.0, "step": 6848 }, { "epoch": 0.8712631980664037, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.083171844482422, "learning_rate": 1e-06, "loss": 0.5151, "mean_token_accuracy": 0.8363993763923645, "num_tokens": 261280330.0, "step": 6849 }, { "epoch": 0.8713904083449943, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.075340270996094, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.873805820941925, "num_tokens": 261318232.0, "step": 6850 }, { "epoch": 0.8715176186235848, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.122478485107422, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8797174692153931, "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": 11.075550079345703, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8660147786140442, "num_tokens": 261387058.0, "step": 6852 }, { "epoch": 0.8717720391807658, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.078941345214844, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8785272836685181, "num_tokens": 261430110.0, "step": 6853 }, { "epoch": 0.8718992494593563, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.067645072937012, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8700763583183289, "num_tokens": 261470707.0, "step": 6854 }, { "epoch": 0.8720264597379468, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.057760238647461, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8705297708511353, "num_tokens": 261507146.0, "step": 6855 }, { "epoch": 0.8721536700165373, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.106407165527344, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8496384024620056, "num_tokens": 261549723.0, "step": 6856 }, { "epoch": 0.8722808802951278, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.091472625732422, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.868677020072937, "num_tokens": 261591030.0, "step": 6857 }, { "epoch": 0.8724080905737184, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.047014236450195, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8726418614387512, "num_tokens": 261623919.0, "step": 6858 }, { "epoch": 0.8725353008523089, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.084189414978027, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8800771236419678, "num_tokens": 261655277.0, "step": 6859 }, { "epoch": 0.8726625111308994, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.109009742736816, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.8511720895767212, "num_tokens": 261697462.0, "step": 6860 }, { "epoch": 0.8727897214094898, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.06658935546875, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8602235913276672, "num_tokens": 261739134.0, "step": 6861 }, { "epoch": 0.8729169316880804, "ewc_loss": 0.0224609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 11.072908401489258, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8670920133590698, "num_tokens": 261776295.0, "step": 6862 }, { "epoch": 0.8730441419666709, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.092314720153809, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8672406673431396, "num_tokens": 261812388.0, "step": 6863 }, { "epoch": 0.8731713522452614, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.052298545837402, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8646901845932007, "num_tokens": 261855537.0, "step": 6864 }, { "epoch": 0.873298562523852, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.038007736206055, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8719216585159302, "num_tokens": 261899922.0, "step": 6865 }, { "epoch": 0.8734257728024425, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.13847827911377, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8751285672187805, "num_tokens": 261934475.0, "step": 6866 }, { "epoch": 0.8735529830810329, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.072148323059082, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8644824028015137, "num_tokens": 261973395.0, "step": 6867 }, { "epoch": 0.8736801933596234, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.083821296691895, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.847143828868866, "num_tokens": 262011644.0, "step": 6868 }, { "epoch": 0.873807403638214, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.077730178833008, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.859093964099884, "num_tokens": 262044038.0, "step": 6869 }, { "epoch": 0.8739346139168045, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.133588790893555, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8627729415893555, "num_tokens": 262081023.0, "step": 6870 }, { "epoch": 0.874061824195395, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.043774604797363, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8573226928710938, "num_tokens": 262123981.0, "step": 6871 }, { "epoch": 0.8741890344739855, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.060129165649414, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8534925580024719, "num_tokens": 262161542.0, "step": 6872 }, { "epoch": 0.874316244752576, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.035727500915527, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8531681299209595, "num_tokens": 262202482.0, "step": 6873 }, { "epoch": 0.8744434550311665, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.0416898727417, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8653092980384827, "num_tokens": 262239499.0, "step": 6874 }, { "epoch": 0.874570665309757, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.13326358795166, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8644953966140747, "num_tokens": 262281324.0, "step": 6875 }, { "epoch": 0.8746978755883476, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.080951690673828, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8549450635910034, "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": 11.078404426574707, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8766474723815918, "num_tokens": 262353264.0, "step": 6877 }, { "epoch": 0.8749522961455286, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.118874549865723, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8613475561141968, "num_tokens": 262393360.0, "step": 6878 }, { "epoch": 0.875079506424119, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.088652610778809, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8742478489875793, "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": 11.054061889648438, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8640643358230591, "num_tokens": 262468828.0, "step": 6880 }, { "epoch": 0.8753339269813001, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.135274887084961, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8707906007766724, "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": 11.060054779052734, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8614251613616943, "num_tokens": 262553137.0, "step": 6882 }, { "epoch": 0.8755883475384811, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.112455368041992, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8866338133811951, "num_tokens": 262593847.0, "step": 6883 }, { "epoch": 0.8757155578170717, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.08378791809082, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8529386520385742, "num_tokens": 262635048.0, "step": 6884 }, { "epoch": 0.8758427680956621, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.162182807922363, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8557819128036499, "num_tokens": 262671200.0, "step": 6885 }, { "epoch": 0.8759699783742526, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.039322853088379, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8651027679443359, "num_tokens": 262708304.0, "step": 6886 }, { "epoch": 0.8760971886528431, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.106047630310059, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8606935739517212, "num_tokens": 262750680.0, "step": 6887 }, { "epoch": 0.8762243989314337, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.0900239944458, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.858655571937561, "num_tokens": 262784845.0, "step": 6888 }, { "epoch": 0.8763516092100242, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.09906005859375, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.863519549369812, "num_tokens": 262822646.0, "step": 6889 }, { "epoch": 0.8764788194886147, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.045292854309082, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8598660230636597, "num_tokens": 262863663.0, "step": 6890 }, { "epoch": 0.8766060297672051, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.109389305114746, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8510779738426208, "num_tokens": 262903726.0, "step": 6891 }, { "epoch": 0.8767332400457957, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.05832290649414, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8532633185386658, "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": 11.138870239257812, "learning_rate": 1e-06, "loss": 0.4883, "mean_token_accuracy": 0.8429158926010132, "num_tokens": 262977639.0, "step": 6893 }, { "epoch": 0.8769876606029767, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.030537605285645, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8660045266151428, "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": 11.111005783081055, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8731058835983276, "num_tokens": 263058618.0, "step": 6895 }, { "epoch": 0.8772420811601578, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.044577598571777, "learning_rate": 1e-06, "loss": 0.4832, "mean_token_accuracy": 0.8470056653022766, "num_tokens": 263098628.0, "step": 6896 }, { "epoch": 0.8773692914387482, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.071276664733887, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8673436641693115, "num_tokens": 263136793.0, "step": 6897 }, { "epoch": 0.8774965017173387, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.093005180358887, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8662979602813721, "num_tokens": 263170996.0, "step": 6898 }, { "epoch": 0.8776237119959293, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.104647636413574, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8536555767059326, "num_tokens": 263199961.0, "step": 6899 }, { "epoch": 0.8777509222745198, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.157170295715332, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8618923425674438, "num_tokens": 263241541.0, "step": 6900 }, { "epoch": 0.8778781325531103, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.03329849243164, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.876062273979187, "num_tokens": 263278778.0, "step": 6901 }, { "epoch": 0.8780053428317008, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.139412879943848, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8661318421363831, "num_tokens": 263310155.0, "step": 6902 }, { "epoch": 0.8781325531102913, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.127614974975586, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8593727350234985, "num_tokens": 263355793.0, "step": 6903 }, { "epoch": 0.8782597633888818, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.04244327545166, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8735862374305725, "num_tokens": 263394446.0, "step": 6904 }, { "epoch": 0.8783869736674723, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.098134994506836, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.866429328918457, "num_tokens": 263438479.0, "step": 6905 }, { "epoch": 0.8785141839460628, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.122908592224121, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8496072292327881, "num_tokens": 263478514.0, "step": 6906 }, { "epoch": 0.8786413942246534, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.096991539001465, "learning_rate": 1e-06, "loss": 0.4987, "mean_token_accuracy": 0.8442018032073975, "num_tokens": 263516734.0, "step": 6907 }, { "epoch": 0.8787686045032439, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.091121673583984, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8537197709083557, "num_tokens": 263560105.0, "step": 6908 }, { "epoch": 0.8788958147818343, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.059181213378906, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.866059422492981, "num_tokens": 263597033.0, "step": 6909 }, { "epoch": 0.8790230250604248, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.101937294006348, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.870888352394104, "num_tokens": 263631106.0, "step": 6910 }, { "epoch": 0.8791502353390154, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.055110931396484, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8573216199874878, "num_tokens": 263668227.0, "step": 6911 }, { "epoch": 0.8792774456176059, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.075247764587402, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8656487464904785, "num_tokens": 263704335.0, "step": 6912 }, { "epoch": 0.8794046558961964, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.054574966430664, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8703235387802124, "num_tokens": 263742532.0, "step": 6913 }, { "epoch": 0.879531866174787, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.039892196655273, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8645451068878174, "num_tokens": 263782321.0, "step": 6914 }, { "epoch": 0.8796590764533775, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.083149909973145, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.87178635597229, "num_tokens": 263818037.0, "step": 6915 }, { "epoch": 0.8797862867319679, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.099498748779297, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8488770723342896, "num_tokens": 263854901.0, "step": 6916 }, { "epoch": 0.8799134970105584, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.06605339050293, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8679805994033813, "num_tokens": 263896241.0, "step": 6917 }, { "epoch": 0.880040707289149, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.094298362731934, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8647245168685913, "num_tokens": 263934673.0, "step": 6918 }, { "epoch": 0.8801679175677395, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.105164527893066, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8548736572265625, "num_tokens": 263973456.0, "step": 6919 }, { "epoch": 0.88029512784633, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.161054611206055, "learning_rate": 1e-06, "loss": 0.486, "mean_token_accuracy": 0.8458684086799622, "num_tokens": 264006050.0, "step": 6920 }, { "epoch": 0.8804223381249205, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.09366226196289, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8736706972122192, "num_tokens": 264042736.0, "step": 6921 }, { "epoch": 0.880549548403511, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.075016975402832, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8547812700271606, "num_tokens": 264084433.0, "step": 6922 }, { "epoch": 0.8806767586821015, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.13388442993164, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8710383176803589, "num_tokens": 264118331.0, "step": 6923 }, { "epoch": 0.880803968960692, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.072342872619629, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8654927015304565, "num_tokens": 264152015.0, "step": 6924 }, { "epoch": 0.8809311792392825, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.126230239868164, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8691282868385315, "num_tokens": 264187085.0, "step": 6925 }, { "epoch": 0.8810583895178731, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.091811180114746, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8592193126678467, "num_tokens": 264225558.0, "step": 6926 }, { "epoch": 0.8811855997964636, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.0566987991333, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8546067476272583, "num_tokens": 264264220.0, "step": 6927 }, { "epoch": 0.881312810075054, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.114727973937988, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8664330244064331, "num_tokens": 264302097.0, "step": 6928 }, { "epoch": 0.8814400203536445, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.043423652648926, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8658115863800049, "num_tokens": 264341966.0, "step": 6929 }, { "epoch": 0.8815672306322351, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.054789543151855, "learning_rate": 1e-06, "loss": 0.508, "mean_token_accuracy": 0.8410723209381104, "num_tokens": 264384298.0, "step": 6930 }, { "epoch": 0.8816944409108256, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.116116523742676, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8638056516647339, "num_tokens": 264422513.0, "step": 6931 }, { "epoch": 0.8818216511894161, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.085201263427734, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8758144378662109, "num_tokens": 264461819.0, "step": 6932 }, { "epoch": 0.8819488614680067, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.069928169250488, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8534363508224487, "num_tokens": 264502618.0, "step": 6933 }, { "epoch": 0.8820760717465971, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.189945220947266, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8559586405754089, "num_tokens": 264534382.0, "step": 6934 }, { "epoch": 0.8822032820251876, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.117470741271973, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8536704778671265, "num_tokens": 264575297.0, "step": 6935 }, { "epoch": 0.8823304923037781, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.080007553100586, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8694127798080444, "num_tokens": 264611822.0, "step": 6936 }, { "epoch": 0.8824577025823687, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.050751686096191, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8589179515838623, "num_tokens": 264657483.0, "step": 6937 }, { "epoch": 0.8825849128609592, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.120798110961914, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8687805533409119, "num_tokens": 264690633.0, "step": 6938 }, { "epoch": 0.8827121231395497, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.082879066467285, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8782210350036621, "num_tokens": 264733268.0, "step": 6939 }, { "epoch": 0.8828393334181401, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.076071739196777, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8566281795501709, "num_tokens": 264774769.0, "step": 6940 }, { "epoch": 0.8829665436967307, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.14346981048584, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8587775826454163, "num_tokens": 264806658.0, "step": 6941 }, { "epoch": 0.8830937539753212, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.083564758300781, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8512250781059265, "num_tokens": 264841054.0, "step": 6942 }, { "epoch": 0.8832209642539117, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.099632263183594, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8581002354621887, "num_tokens": 264880241.0, "step": 6943 }, { "epoch": 0.8833481745325023, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.032326698303223, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8655222654342651, "num_tokens": 264922449.0, "step": 6944 }, { "epoch": 0.8834753848110928, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.114080429077148, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8630259037017822, "num_tokens": 264961605.0, "step": 6945 }, { "epoch": 0.8836025950896832, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.062264442443848, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8708403706550598, "num_tokens": 264992924.0, "step": 6946 }, { "epoch": 0.8837298053682737, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.123990058898926, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8624327182769775, "num_tokens": 265031849.0, "step": 6947 }, { "epoch": 0.8838570156468643, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.076279640197754, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8753029704093933, "num_tokens": 265069714.0, "step": 6948 }, { "epoch": 0.8839842259254548, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.098234176635742, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8586390614509583, "num_tokens": 265103325.0, "step": 6949 }, { "epoch": 0.8841114362040453, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.040936470031738, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.8419796824455261, "num_tokens": 265139133.0, "step": 6950 }, { "epoch": 0.8842386464826358, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.07614803314209, "learning_rate": 1e-06, "loss": 0.5024, "mean_token_accuracy": 0.8388911485671997, "num_tokens": 265183536.0, "step": 6951 }, { "epoch": 0.8843658567612263, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.101508140563965, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8548135161399841, "num_tokens": 265226288.0, "step": 6952 }, { "epoch": 0.8844930670398168, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.101689338684082, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8629595041275024, "num_tokens": 265265846.0, "step": 6953 }, { "epoch": 0.8846202773184073, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.10107135772705, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8766405582427979, "num_tokens": 265300612.0, "step": 6954 }, { "epoch": 0.8847474875969978, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.115632057189941, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8718355298042297, "num_tokens": 265341440.0, "step": 6955 }, { "epoch": 0.8848746978755884, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.053702354431152, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8537429571151733, "num_tokens": 265384854.0, "step": 6956 }, { "epoch": 0.8850019081541789, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.066650390625, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8565157651901245, "num_tokens": 265420611.0, "step": 6957 }, { "epoch": 0.8851291184327693, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.12928581237793, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8612213134765625, "num_tokens": 265457720.0, "step": 6958 }, { "epoch": 0.8852563287113598, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.084561347961426, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8709359169006348, "num_tokens": 265493434.0, "step": 6959 }, { "epoch": 0.8853835389899504, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.078267097473145, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8650146722793579, "num_tokens": 265529015.0, "step": 6960 }, { "epoch": 0.8855107492685409, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.1138277053833, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8477652668952942, "num_tokens": 265568698.0, "step": 6961 }, { "epoch": 0.8856379595471314, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.060982704162598, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8633955121040344, "num_tokens": 265605631.0, "step": 6962 }, { "epoch": 0.885765169825722, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.12105655670166, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8647910356521606, "num_tokens": 265647071.0, "step": 6963 }, { "epoch": 0.8858923801043125, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.058429718017578, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8632711172103882, "num_tokens": 265688680.0, "step": 6964 }, { "epoch": 0.8860195903829029, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.11159896850586, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8690158128738403, "num_tokens": 265727707.0, "step": 6965 }, { "epoch": 0.8861468006614934, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.090982437133789, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8604733347892761, "num_tokens": 265765860.0, "step": 6966 }, { "epoch": 0.886274010940084, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.14322280883789, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8671407103538513, "num_tokens": 265802988.0, "step": 6967 }, { "epoch": 0.8864012212186745, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.078340530395508, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8579155802726746, "num_tokens": 265840614.0, "step": 6968 }, { "epoch": 0.886528431497265, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.128900527954102, "learning_rate": 1e-06, "loss": 0.5042, "mean_token_accuracy": 0.8364053964614868, "num_tokens": 265883032.0, "step": 6969 }, { "epoch": 0.8866556417758555, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.066144943237305, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8747187852859497, "num_tokens": 265918309.0, "step": 6970 }, { "epoch": 0.886782852054446, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.130460739135742, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8651320934295654, "num_tokens": 265956406.0, "step": 6971 }, { "epoch": 0.8869100623330365, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.096793174743652, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8715196251869202, "num_tokens": 265988413.0, "step": 6972 }, { "epoch": 0.887037272611627, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.109027862548828, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8619862794876099, "num_tokens": 266022036.0, "step": 6973 }, { "epoch": 0.8871644828902175, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.149171829223633, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8622488975524902, "num_tokens": 266053735.0, "step": 6974 }, { "epoch": 0.8872916931688081, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.068333625793457, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.877192497253418, "num_tokens": 266094198.0, "step": 6975 }, { "epoch": 0.8874189034473986, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.145201683044434, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8717976808547974, "num_tokens": 266132213.0, "step": 6976 }, { "epoch": 0.887546113725989, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.103243827819824, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.863089919090271, "num_tokens": 266168001.0, "step": 6977 }, { "epoch": 0.8876733240045795, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.106362342834473, "learning_rate": 1e-06, "loss": 0.505, "mean_token_accuracy": 0.8348251581192017, "num_tokens": 266208771.0, "step": 6978 }, { "epoch": 0.8878005342831701, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.107864379882812, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8694298267364502, "num_tokens": 266248090.0, "step": 6979 }, { "epoch": 0.8879277445617606, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.153910636901855, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8522318005561829, "num_tokens": 266283748.0, "step": 6980 }, { "epoch": 0.8880549548403511, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.094343185424805, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.869727611541748, "num_tokens": 266316613.0, "step": 6981 }, { "epoch": 0.8881821651189417, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.120233535766602, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8653354644775391, "num_tokens": 266357515.0, "step": 6982 }, { "epoch": 0.8883093753975321, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.091140747070312, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8750537037849426, "num_tokens": 266393593.0, "step": 6983 }, { "epoch": 0.8884365856761226, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.122017860412598, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.87372887134552, "num_tokens": 266436104.0, "step": 6984 }, { "epoch": 0.8885637959547131, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.116976737976074, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8756251335144043, "num_tokens": 266472809.0, "step": 6985 }, { "epoch": 0.8886910062333037, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.05809211730957, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8733241558074951, "num_tokens": 266512726.0, "step": 6986 }, { "epoch": 0.8888182165118942, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.146872520446777, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8521368503570557, "num_tokens": 266546098.0, "step": 6987 }, { "epoch": 0.8889454267904847, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.088663101196289, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8540007472038269, "num_tokens": 266588478.0, "step": 6988 }, { "epoch": 0.8890726370690751, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.095351219177246, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8525703549385071, "num_tokens": 266628975.0, "step": 6989 }, { "epoch": 0.8891998473476657, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.031973838806152, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8719587326049805, "num_tokens": 266667533.0, "step": 6990 }, { "epoch": 0.8893270576262562, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.13709831237793, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8439819812774658, "num_tokens": 266706990.0, "step": 6991 }, { "epoch": 0.8894542679048467, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.03893756866455, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.851550281047821, "num_tokens": 266747314.0, "step": 6992 }, { "epoch": 0.8895814781834372, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.072461128234863, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8537983894348145, "num_tokens": 266791031.0, "step": 6993 }, { "epoch": 0.8897086884620278, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.098912239074707, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.86289381980896, "num_tokens": 266823236.0, "step": 6994 }, { "epoch": 0.8898358987406182, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.070037841796875, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8613544702529907, "num_tokens": 266867993.0, "step": 6995 }, { "epoch": 0.8899631090192087, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.061609268188477, "learning_rate": 1e-06, "loss": 0.4807, "mean_token_accuracy": 0.8448770642280579, "num_tokens": 266908573.0, "step": 6996 }, { "epoch": 0.8900903192977992, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.10783863067627, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8606152534484863, "num_tokens": 266951167.0, "step": 6997 }, { "epoch": 0.8902175295763898, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.070819854736328, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8621913194656372, "num_tokens": 266988976.0, "step": 6998 }, { "epoch": 0.8903447398549803, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.124282836914062, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8621001839637756, "num_tokens": 267022450.0, "step": 6999 }, { "epoch": 0.8904719501335708, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.12637710571289, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.856942892074585, "num_tokens": 267059933.0, "step": 7000 }, { "epoch": 0.8905991604121613, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.120450019836426, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8542163372039795, "num_tokens": 267092518.0, "step": 7001 }, { "epoch": 0.8907263706907518, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.13314151763916, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8578342795372009, "num_tokens": 267130008.0, "step": 7002 }, { "epoch": 0.8908535809693423, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.079360008239746, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8659560680389404, "num_tokens": 267171039.0, "step": 7003 }, { "epoch": 0.8909807912479328, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.118905067443848, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8656269311904907, "num_tokens": 267211197.0, "step": 7004 }, { "epoch": 0.8911080015265234, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.074703216552734, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8677825927734375, "num_tokens": 267247471.0, "step": 7005 }, { "epoch": 0.8912352118051139, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.117115020751953, "learning_rate": 1e-06, "loss": 0.517, "mean_token_accuracy": 0.8354832530021667, "num_tokens": 267287276.0, "step": 7006 }, { "epoch": 0.8913624220837043, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.114789009094238, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8598113059997559, "num_tokens": 267323827.0, "step": 7007 }, { "epoch": 0.8914896323622948, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.047368049621582, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8712590336799622, "num_tokens": 267364053.0, "step": 7008 }, { "epoch": 0.8916168426408854, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.082768440246582, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8460865020751953, "num_tokens": 267404393.0, "step": 7009 }, { "epoch": 0.8917440529194759, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.116466522216797, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8629461526870728, "num_tokens": 267444589.0, "step": 7010 }, { "epoch": 0.8918712631980664, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.142891883850098, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8634753227233887, "num_tokens": 267484460.0, "step": 7011 }, { "epoch": 0.891998473476657, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.100631713867188, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8550752401351929, "num_tokens": 267517390.0, "step": 7012 }, { "epoch": 0.8921256837552475, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.147985458374023, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8628748655319214, "num_tokens": 267555813.0, "step": 7013 }, { "epoch": 0.8922528940338379, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.085542678833008, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8604645729064941, "num_tokens": 267595818.0, "step": 7014 }, { "epoch": 0.8923801043124284, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.169807434082031, "learning_rate": 1e-06, "loss": 0.508, "mean_token_accuracy": 0.836784839630127, "num_tokens": 267633106.0, "step": 7015 }, { "epoch": 0.892507314591019, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.14079761505127, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8647414445877075, "num_tokens": 267670166.0, "step": 7016 }, { "epoch": 0.8926345248696095, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.096760749816895, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8589447140693665, "num_tokens": 267707284.0, "step": 7017 }, { "epoch": 0.8927617351482, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 11.135746955871582, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.862231969833374, "num_tokens": 267742349.0, "step": 7018 }, { "epoch": 0.8928889454267905, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.158284187316895, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.867194414138794, "num_tokens": 267776003.0, "step": 7019 }, { "epoch": 0.893016155705381, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.094575881958008, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8577919006347656, "num_tokens": 267815887.0, "step": 7020 }, { "epoch": 0.8931433659839715, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.203042984008789, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8725574016571045, "num_tokens": 267855874.0, "step": 7021 }, { "epoch": 0.893270576262562, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.097699165344238, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8724253177642822, "num_tokens": 267897901.0, "step": 7022 }, { "epoch": 0.8933977865411525, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.159384727478027, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.872833251953125, "num_tokens": 267938967.0, "step": 7023 }, { "epoch": 0.8935249968197431, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.13151741027832, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8594995737075806, "num_tokens": 267975196.0, "step": 7024 }, { "epoch": 0.8936522070983336, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.139211654663086, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8756108283996582, "num_tokens": 268014407.0, "step": 7025 }, { "epoch": 0.893779417376924, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.117907524108887, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8777177333831787, "num_tokens": 268043892.0, "step": 7026 }, { "epoch": 0.8939066276555145, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.070310592651367, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8623054027557373, "num_tokens": 268079651.0, "step": 7027 }, { "epoch": 0.8940338379341051, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.135499000549316, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8652281761169434, "num_tokens": 268114021.0, "step": 7028 }, { "epoch": 0.8941610482126956, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.063645362854004, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8613941073417664, "num_tokens": 268155315.0, "step": 7029 }, { "epoch": 0.8942882584912861, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.13888168334961, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8597780466079712, "num_tokens": 268199312.0, "step": 7030 }, { "epoch": 0.8944154687698767, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.072734832763672, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8692220449447632, "num_tokens": 268236955.0, "step": 7031 }, { "epoch": 0.8945426790484671, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.139347076416016, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8591312170028687, "num_tokens": 268272183.0, "step": 7032 }, { "epoch": 0.8946698893270576, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.12114429473877, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8532193303108215, "num_tokens": 268305407.0, "step": 7033 }, { "epoch": 0.8947970996056481, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.146202087402344, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8738949298858643, "num_tokens": 268340443.0, "step": 7034 }, { "epoch": 0.8949243098842387, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.059659004211426, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8812509775161743, "num_tokens": 268375597.0, "step": 7035 }, { "epoch": 0.8950515201628292, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.093658447265625, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8560420870780945, "num_tokens": 268412872.0, "step": 7036 }, { "epoch": 0.8951787304414197, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.064050674438477, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8635678291320801, "num_tokens": 268452021.0, "step": 7037 }, { "epoch": 0.8953059407200101, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.1013765335083, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8723281621932983, "num_tokens": 268486354.0, "step": 7038 }, { "epoch": 0.8954331509986007, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.085768699645996, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8562954664230347, "num_tokens": 268519918.0, "step": 7039 }, { "epoch": 0.8955603612771912, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.137787818908691, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8545631170272827, "num_tokens": 268554259.0, "step": 7040 }, { "epoch": 0.8956875715557817, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.149801254272461, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8571345806121826, "num_tokens": 268595785.0, "step": 7041 }, { "epoch": 0.8958147818343722, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.090629577636719, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.847406804561615, "num_tokens": 268640805.0, "step": 7042 }, { "epoch": 0.8959419921129628, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.105761528015137, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8682471513748169, "num_tokens": 268682446.0, "step": 7043 }, { "epoch": 0.8960692023915532, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.133360862731934, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8707612752914429, "num_tokens": 268720262.0, "step": 7044 }, { "epoch": 0.8961964126701437, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.116100311279297, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8490862846374512, "num_tokens": 268760519.0, "step": 7045 }, { "epoch": 0.8963236229487342, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.118877410888672, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8563916683197021, "num_tokens": 268796621.0, "step": 7046 }, { "epoch": 0.8964508332273248, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.15897274017334, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8711211681365967, "num_tokens": 268830527.0, "step": 7047 }, { "epoch": 0.8965780435059153, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.136608123779297, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8666397929191589, "num_tokens": 268862112.0, "step": 7048 }, { "epoch": 0.8967052537845058, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.122182846069336, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8605325222015381, "num_tokens": 268901089.0, "step": 7049 }, { "epoch": 0.8968324640630962, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.1417875289917, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8647900819778442, "num_tokens": 268942766.0, "step": 7050 }, { "epoch": 0.8969596743416868, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.093098640441895, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8696107864379883, "num_tokens": 268978642.0, "step": 7051 }, { "epoch": 0.8970868846202773, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.136417388916016, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8684291839599609, "num_tokens": 269016048.0, "step": 7052 }, { "epoch": 0.8972140948988678, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.09508991241455, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8632497787475586, "num_tokens": 269054381.0, "step": 7053 }, { "epoch": 0.8973413051774584, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.107990264892578, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8662471771240234, "num_tokens": 269090672.0, "step": 7054 }, { "epoch": 0.8974685154560489, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.089739799499512, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8663999438285828, "num_tokens": 269130231.0, "step": 7055 }, { "epoch": 0.8975957257346393, "ewc_loss": 0.022705078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.09470272064209, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8676823973655701, "num_tokens": 269166282.0, "step": 7056 }, { "epoch": 0.8977229360132298, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.114487648010254, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8659123182296753, "num_tokens": 269205326.0, "step": 7057 }, { "epoch": 0.8978501462918204, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.125969886779785, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8647457957267761, "num_tokens": 269243903.0, "step": 7058 }, { "epoch": 0.8979773565704109, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.067584037780762, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8572028875350952, "num_tokens": 269288330.0, "step": 7059 }, { "epoch": 0.8981045668490014, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.096025466918945, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.853073239326477, "num_tokens": 269328731.0, "step": 7060 }, { "epoch": 0.898231777127592, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.18177604675293, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8636177182197571, "num_tokens": 269369576.0, "step": 7061 }, { "epoch": 0.8983589874061825, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.118988037109375, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8616125583648682, "num_tokens": 269407734.0, "step": 7062 }, { "epoch": 0.8984861976847729, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.125229835510254, "learning_rate": 1e-06, "loss": 0.4905, "mean_token_accuracy": 0.8428941965103149, "num_tokens": 269444042.0, "step": 7063 }, { "epoch": 0.8986134079633634, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.134636878967285, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8860563039779663, "num_tokens": 269480979.0, "step": 7064 }, { "epoch": 0.898740618241954, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.142006874084473, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8642645478248596, "num_tokens": 269515594.0, "step": 7065 }, { "epoch": 0.8988678285205445, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.067989349365234, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.865799069404602, "num_tokens": 269556974.0, "step": 7066 }, { "epoch": 0.898995038799135, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.196489334106445, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8561254739761353, "num_tokens": 269591088.0, "step": 7067 }, { "epoch": 0.8991222490777255, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.123520851135254, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8717002868652344, "num_tokens": 269627921.0, "step": 7068 }, { "epoch": 0.899249459356316, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.12987995147705, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8473320007324219, "num_tokens": 269666586.0, "step": 7069 }, { "epoch": 0.8993766696349065, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.152700424194336, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8658527731895447, "num_tokens": 269701205.0, "step": 7070 }, { "epoch": 0.899503879913497, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.109068870544434, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8754757642745972, "num_tokens": 269739982.0, "step": 7071 }, { "epoch": 0.8996310901920875, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.138114929199219, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8641231656074524, "num_tokens": 269773048.0, "step": 7072 }, { "epoch": 0.8997583004706781, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.1461181640625, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8622533082962036, "num_tokens": 269811343.0, "step": 7073 }, { "epoch": 0.8998855107492686, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.081317901611328, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8678086996078491, "num_tokens": 269851601.0, "step": 7074 }, { "epoch": 0.900012721027859, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.140243530273438, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8718337416648865, "num_tokens": 269891396.0, "step": 7075 }, { "epoch": 0.9001399313064495, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.127641677856445, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8628722429275513, "num_tokens": 269930292.0, "step": 7076 }, { "epoch": 0.9002671415850401, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.15590763092041, "learning_rate": 1e-06, "loss": 0.4949, "mean_token_accuracy": 0.836643636226654, "num_tokens": 269968464.0, "step": 7077 }, { "epoch": 0.9003943518636306, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.074125289916992, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8505454659461975, "num_tokens": 270011569.0, "step": 7078 }, { "epoch": 0.9005215621422211, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.120233535766602, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8594150543212891, "num_tokens": 270050608.0, "step": 7079 }, { "epoch": 0.9006487724208116, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.068824768066406, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8489570021629333, "num_tokens": 270096100.0, "step": 7080 }, { "epoch": 0.9007759826994021, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.085203170776367, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8739268779754639, "num_tokens": 270140915.0, "step": 7081 }, { "epoch": 0.9009031929779926, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.124567985534668, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8628371953964233, "num_tokens": 270184233.0, "step": 7082 }, { "epoch": 0.9010304032565831, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.090051651000977, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8616354465484619, "num_tokens": 270223889.0, "step": 7083 }, { "epoch": 0.9011576135351737, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.148527145385742, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8683570623397827, "num_tokens": 270266357.0, "step": 7084 }, { "epoch": 0.9012848238137642, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.115234375, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8713698387145996, "num_tokens": 270298252.0, "step": 7085 }, { "epoch": 0.9014120340923547, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.122979164123535, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8564881086349487, "num_tokens": 270343187.0, "step": 7086 }, { "epoch": 0.9015392443709451, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.113334655761719, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.852616548538208, "num_tokens": 270385094.0, "step": 7087 }, { "epoch": 0.9016664546495357, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.103717803955078, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8500843048095703, "num_tokens": 270421772.0, "step": 7088 }, { "epoch": 0.9017936649281262, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.063297271728516, "learning_rate": 1e-06, "loss": 0.4882, "mean_token_accuracy": 0.8448201417922974, "num_tokens": 270465033.0, "step": 7089 }, { "epoch": 0.9019208752067167, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.1396484375, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8847922682762146, "num_tokens": 270501370.0, "step": 7090 }, { "epoch": 0.9020480854853072, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.119226455688477, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.867759108543396, "num_tokens": 270543571.0, "step": 7091 }, { "epoch": 0.9021752957638978, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.129924774169922, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.873274564743042, "num_tokens": 270580635.0, "step": 7092 }, { "epoch": 0.9023025060424882, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.065122604370117, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8760051727294922, "num_tokens": 270616534.0, "step": 7093 }, { "epoch": 0.9024297163210787, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.153583526611328, "learning_rate": 1e-06, "loss": 0.492, "mean_token_accuracy": 0.8439524173736572, "num_tokens": 270653998.0, "step": 7094 }, { "epoch": 0.9025569265996692, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.144233703613281, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8639264702796936, "num_tokens": 270688023.0, "step": 7095 }, { "epoch": 0.9026841368782598, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.164323806762695, "learning_rate": 1e-06, "loss": 0.4872, "mean_token_accuracy": 0.8453388214111328, "num_tokens": 270721298.0, "step": 7096 }, { "epoch": 0.9028113471568503, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.104194641113281, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8685340881347656, "num_tokens": 270762182.0, "step": 7097 }, { "epoch": 0.9029385574354408, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.13530445098877, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.878582239151001, "num_tokens": 270799577.0, "step": 7098 }, { "epoch": 0.9030657677140312, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.14954662322998, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8638466596603394, "num_tokens": 270840218.0, "step": 7099 }, { "epoch": 0.9031929779926218, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.108955383300781, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8844983577728271, "num_tokens": 270872982.0, "step": 7100 }, { "epoch": 0.9033201882712123, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.206876754760742, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8682472109794617, "num_tokens": 270908381.0, "step": 7101 }, { "epoch": 0.9034473985498028, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.036941528320312, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8615415692329407, "num_tokens": 270949217.0, "step": 7102 }, { "epoch": 0.9035746088283934, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.198952674865723, "learning_rate": 1e-06, "loss": 0.477, "mean_token_accuracy": 0.8472789525985718, "num_tokens": 270989667.0, "step": 7103 }, { "epoch": 0.9037018191069839, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.071272850036621, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8532938957214355, "num_tokens": 271030395.0, "step": 7104 }, { "epoch": 0.9038290293855743, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.171025276184082, "learning_rate": 1e-06, "loss": 0.4913, "mean_token_accuracy": 0.8458853363990784, "num_tokens": 271068239.0, "step": 7105 }, { "epoch": 0.9039562396641648, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.15079116821289, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.856250524520874, "num_tokens": 271110294.0, "step": 7106 }, { "epoch": 0.9040834499427554, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.11523151397705, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.880472719669342, "num_tokens": 271143288.0, "step": 7107 }, { "epoch": 0.9042106602213459, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.12749195098877, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8708442449569702, "num_tokens": 271181183.0, "step": 7108 }, { "epoch": 0.9043378704999364, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.120606422424316, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8625051975250244, "num_tokens": 271216201.0, "step": 7109 }, { "epoch": 0.9044650807785269, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.132729530334473, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8695129752159119, "num_tokens": 271250158.0, "step": 7110 }, { "epoch": 0.9045922910571175, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.184483528137207, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8760592341423035, "num_tokens": 271289994.0, "step": 7111 }, { "epoch": 0.9047195013357079, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.167040824890137, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.869867742061615, "num_tokens": 271325782.0, "step": 7112 }, { "epoch": 0.9048467116142984, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.150771141052246, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8502270579338074, "num_tokens": 271361604.0, "step": 7113 }, { "epoch": 0.9049739218928889, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.209460258483887, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8725793361663818, "num_tokens": 271398520.0, "step": 7114 }, { "epoch": 0.9051011321714795, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.114083290100098, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8577246069908142, "num_tokens": 271445138.0, "step": 7115 }, { "epoch": 0.90522834245007, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.170166015625, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8690940141677856, "num_tokens": 271479147.0, "step": 7116 }, { "epoch": 0.9053555527286605, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.124058723449707, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8656433820724487, "num_tokens": 271512477.0, "step": 7117 }, { "epoch": 0.905482763007251, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.181899070739746, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8501614332199097, "num_tokens": 271552355.0, "step": 7118 }, { "epoch": 0.9056099732858415, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.087013244628906, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8596422076225281, "num_tokens": 271595825.0, "step": 7119 }, { "epoch": 0.905737183564432, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.250350952148438, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8509988784790039, "num_tokens": 271627384.0, "step": 7120 }, { "epoch": 0.9058643938430225, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.169289588928223, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8728777170181274, "num_tokens": 271666711.0, "step": 7121 }, { "epoch": 0.9059916041216131, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.121585845947266, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8696435689926147, "num_tokens": 271703145.0, "step": 7122 }, { "epoch": 0.9061188144002036, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.18699836730957, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8519788980484009, "num_tokens": 271743131.0, "step": 7123 }, { "epoch": 0.906246024678794, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.161733627319336, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8668875694274902, "num_tokens": 271778245.0, "step": 7124 }, { "epoch": 0.9063732349573845, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.186256408691406, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8554284572601318, "num_tokens": 271817279.0, "step": 7125 }, { "epoch": 0.9065004452359751, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.11800479888916, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8706884384155273, "num_tokens": 271857136.0, "step": 7126 }, { "epoch": 0.9066276555145656, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.221283912658691, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8718354105949402, "num_tokens": 271891651.0, "step": 7127 }, { "epoch": 0.9067548657931561, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.122950553894043, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8689330220222473, "num_tokens": 271925805.0, "step": 7128 }, { "epoch": 0.9068820760717466, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.137880325317383, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8676165342330933, "num_tokens": 271964954.0, "step": 7129 }, { "epoch": 0.9070092863503371, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.191732406616211, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8631476163864136, "num_tokens": 272000331.0, "step": 7130 }, { "epoch": 0.9071364966289276, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.122604370117188, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8707157373428345, "num_tokens": 272037853.0, "step": 7131 }, { "epoch": 0.9072637069075181, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.128854751586914, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8732700347900391, "num_tokens": 272077663.0, "step": 7132 }, { "epoch": 0.9073909171861086, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.113937377929688, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8589142560958862, "num_tokens": 272116942.0, "step": 7133 }, { "epoch": 0.9075181274646992, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.18530559539795, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8574241995811462, "num_tokens": 272151319.0, "step": 7134 }, { "epoch": 0.9076453377432897, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.153106689453125, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8510726094245911, "num_tokens": 272190817.0, "step": 7135 }, { "epoch": 0.9077725480218801, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.118152618408203, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8575047850608826, "num_tokens": 272230461.0, "step": 7136 }, { "epoch": 0.9078997583004706, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.128946304321289, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8578236103057861, "num_tokens": 272268510.0, "step": 7137 }, { "epoch": 0.9080269685790612, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.150497436523438, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8636748790740967, "num_tokens": 272312430.0, "step": 7138 }, { "epoch": 0.9081541788576517, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.120321273803711, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8623492121696472, "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": 11.197529792785645, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8570192456245422, "num_tokens": 272378902.0, "step": 7140 }, { "epoch": 0.9084085994148328, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.14823055267334, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8775053024291992, "num_tokens": 272421257.0, "step": 7141 }, { "epoch": 0.9085358096934232, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.077239036560059, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8626338839530945, "num_tokens": 272459800.0, "step": 7142 }, { "epoch": 0.9086630199720137, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.172682762145996, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8821089267730713, "num_tokens": 272501245.0, "step": 7143 }, { "epoch": 0.9087902302506042, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.166082382202148, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8653080463409424, "num_tokens": 272535873.0, "step": 7144 }, { "epoch": 0.9089174405291948, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.151712417602539, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8612791895866394, "num_tokens": 272575213.0, "step": 7145 }, { "epoch": 0.9090446508077853, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.116927146911621, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.864804744720459, "num_tokens": 272611960.0, "step": 7146 }, { "epoch": 0.9091718610863758, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.170442581176758, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8643028140068054, "num_tokens": 272650895.0, "step": 7147 }, { "epoch": 0.9092990713649662, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.110980987548828, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.863929033279419, "num_tokens": 272690617.0, "step": 7148 }, { "epoch": 0.9094262816435568, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.120561599731445, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8551688194274902, "num_tokens": 272732526.0, "step": 7149 }, { "epoch": 0.9095534919221473, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.160575866699219, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8611003756523132, "num_tokens": 272769882.0, "step": 7150 }, { "epoch": 0.9096807022007378, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.130110740661621, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8671491146087646, "num_tokens": 272809614.0, "step": 7151 }, { "epoch": 0.9098079124793284, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.113394737243652, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8556725978851318, "num_tokens": 272844819.0, "step": 7152 }, { "epoch": 0.9099351227579189, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.11633014678955, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8689754009246826, "num_tokens": 272882798.0, "step": 7153 }, { "epoch": 0.9100623330365093, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.12972640991211, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8663480281829834, "num_tokens": 272917472.0, "step": 7154 }, { "epoch": 0.9101895433150998, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.175224304199219, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8527944087982178, "num_tokens": 272959947.0, "step": 7155 }, { "epoch": 0.9103167535936904, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.116519927978516, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.846467137336731, "num_tokens": 272998194.0, "step": 7156 }, { "epoch": 0.9104439638722809, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.181884765625, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8806489109992981, "num_tokens": 273034688.0, "step": 7157 }, { "epoch": 0.9105711741508714, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.141895294189453, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8735995292663574, "num_tokens": 273071575.0, "step": 7158 }, { "epoch": 0.9106983844294619, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.196334838867188, "learning_rate": 1e-06, "loss": 0.5073, "mean_token_accuracy": 0.8381352424621582, "num_tokens": 273106106.0, "step": 7159 }, { "epoch": 0.9108255947080524, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.198528289794922, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8582968711853027, "num_tokens": 273142959.0, "step": 7160 }, { "epoch": 0.9109528049866429, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.205747604370117, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8488974571228027, "num_tokens": 273182781.0, "step": 7161 }, { "epoch": 0.9110800152652334, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.161629676818848, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8695805668830872, "num_tokens": 273225738.0, "step": 7162 }, { "epoch": 0.9112072255438239, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.137749671936035, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.870861828327179, "num_tokens": 273269166.0, "step": 7163 }, { "epoch": 0.9113344358224145, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.167899131774902, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8581662178039551, "num_tokens": 273309093.0, "step": 7164 }, { "epoch": 0.911461646101005, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.160008430480957, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8685164451599121, "num_tokens": 273346857.0, "step": 7165 }, { "epoch": 0.9115888563795955, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.180432319641113, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8582306504249573, "num_tokens": 273385720.0, "step": 7166 }, { "epoch": 0.9117160666581859, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.094670295715332, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8645017147064209, "num_tokens": 273431079.0, "step": 7167 }, { "epoch": 0.9118432769367765, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.111934661865234, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8647422790527344, "num_tokens": 273477316.0, "step": 7168 }, { "epoch": 0.911970487215367, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.187152862548828, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8671661019325256, "num_tokens": 273512831.0, "step": 7169 }, { "epoch": 0.9120976974939575, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.112360000610352, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8577185273170471, "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": 11.176830291748047, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8658589124679565, "num_tokens": 273583053.0, "step": 7171 }, { "epoch": 0.9123521180511386, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.19163703918457, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8532368540763855, "num_tokens": 273617953.0, "step": 7172 }, { "epoch": 0.912479328329729, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.104341506958008, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8496777415275574, "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": 11.160682678222656, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8536540269851685, "num_tokens": 273682666.0, "step": 7174 }, { "epoch": 0.9127337488869101, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.148025512695312, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8822823762893677, "num_tokens": 273718641.0, "step": 7175 }, { "epoch": 0.9128609591655006, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.107036590576172, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.872943639755249, "num_tokens": 273760502.0, "step": 7176 }, { "epoch": 0.9129881694440911, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.195962905883789, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8581874370574951, "num_tokens": 273795068.0, "step": 7177 }, { "epoch": 0.9131153797226816, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.1803560256958, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8450953960418701, "num_tokens": 273825099.0, "step": 7178 }, { "epoch": 0.9132425900012721, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.142463684082031, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8696460723876953, "num_tokens": 273862566.0, "step": 7179 }, { "epoch": 0.9133698002798626, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.155302047729492, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8643282651901245, "num_tokens": 273896107.0, "step": 7180 }, { "epoch": 0.9134970105584531, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.171746253967285, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8684567809104919, "num_tokens": 273934066.0, "step": 7181 }, { "epoch": 0.9136242208370436, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.157665252685547, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8618847131729126, "num_tokens": 273971843.0, "step": 7182 }, { "epoch": 0.9137514311156342, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.236282348632812, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.856780469417572, "num_tokens": 274001646.0, "step": 7183 }, { "epoch": 0.9138786413942247, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.156637191772461, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8487125635147095, "num_tokens": 274042656.0, "step": 7184 }, { "epoch": 0.9140058516728151, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.134008407592773, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8502241373062134, "num_tokens": 274084609.0, "step": 7185 }, { "epoch": 0.9141330619514056, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.198771476745605, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8662671446800232, "num_tokens": 274127503.0, "step": 7186 }, { "epoch": 0.9142602722299962, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.11478328704834, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8678063154220581, "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": 11.198441505432129, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8628692626953125, "num_tokens": 274198104.0, "step": 7188 }, { "epoch": 0.9145146927871772, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.211848258972168, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8528478145599365, "num_tokens": 274233487.0, "step": 7189 }, { "epoch": 0.9146419030657678, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.211005210876465, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8670945167541504, "num_tokens": 274267317.0, "step": 7190 }, { "epoch": 0.9147691133443582, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.128085136413574, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8567579984664917, "num_tokens": 274304169.0, "step": 7191 }, { "epoch": 0.9148963236229487, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.217455863952637, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8641432523727417, "num_tokens": 274343265.0, "step": 7192 }, { "epoch": 0.9150235339015392, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.175960540771484, "learning_rate": 1e-06, "loss": 0.4728, "mean_token_accuracy": 0.8466167449951172, "num_tokens": 274380756.0, "step": 7193 }, { "epoch": 0.9151507441801298, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.136841773986816, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8730832934379578, "num_tokens": 274422035.0, "step": 7194 }, { "epoch": 0.9152779544587203, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.116153717041016, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8527735471725464, "num_tokens": 274460875.0, "step": 7195 }, { "epoch": 0.9154051647373108, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.245205879211426, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8657404184341431, "num_tokens": 274495807.0, "step": 7196 }, { "epoch": 0.9155323750159012, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.166070938110352, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8714731931686401, "num_tokens": 274533098.0, "step": 7197 }, { "epoch": 0.9156595852944918, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.2310791015625, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8589376211166382, "num_tokens": 274565282.0, "step": 7198 }, { "epoch": 0.9157867955730823, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.138912200927734, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8661591410636902, "num_tokens": 274605401.0, "step": 7199 }, { "epoch": 0.9159140058516728, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.217245101928711, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8577362298965454, "num_tokens": 274645018.0, "step": 7200 }, { "epoch": 0.9160412161302633, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.173173904418945, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8726437091827393, "num_tokens": 274681541.0, "step": 7201 }, { "epoch": 0.9161684264088539, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.219195365905762, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8678609132766724, "num_tokens": 274716883.0, "step": 7202 }, { "epoch": 0.9162956366874443, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.176809310913086, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8575060367584229, "num_tokens": 274753038.0, "step": 7203 }, { "epoch": 0.9164228469660348, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.21275520324707, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8695343732833862, "num_tokens": 274790760.0, "step": 7204 }, { "epoch": 0.9165500572446253, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.170044898986816, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8555798530578613, "num_tokens": 274830872.0, "step": 7205 }, { "epoch": 0.9166772675232159, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.144819259643555, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8501227498054504, "num_tokens": 274869832.0, "step": 7206 }, { "epoch": 0.9168044778018064, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.22072982788086, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8555580973625183, "num_tokens": 274903853.0, "step": 7207 }, { "epoch": 0.9169316880803969, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.165861129760742, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8626607656478882, "num_tokens": 274942911.0, "step": 7208 }, { "epoch": 0.9170588983589874, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.144545555114746, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8701328635215759, "num_tokens": 274986427.0, "step": 7209 }, { "epoch": 0.9171861086375779, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.211864471435547, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8664100170135498, "num_tokens": 275021557.0, "step": 7210 }, { "epoch": 0.9173133189161684, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.190704345703125, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8518384695053101, "num_tokens": 275061308.0, "step": 7211 }, { "epoch": 0.9174405291947589, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.191041946411133, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8561357855796814, "num_tokens": 275101497.0, "step": 7212 }, { "epoch": 0.9175677394733495, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.185332298278809, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.851561427116394, "num_tokens": 275138317.0, "step": 7213 }, { "epoch": 0.91769494975194, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.170437812805176, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8623580932617188, "num_tokens": 275180119.0, "step": 7214 }, { "epoch": 0.9178221600305305, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.185890197753906, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.86104416847229, "num_tokens": 275212083.0, "step": 7215 }, { "epoch": 0.9179493703091209, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.15499496459961, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8528580069541931, "num_tokens": 275247872.0, "step": 7216 }, { "epoch": 0.9180765805877115, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.113879203796387, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8671073913574219, "num_tokens": 275287022.0, "step": 7217 }, { "epoch": 0.918203790866302, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.190881729125977, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8654986619949341, "num_tokens": 275322259.0, "step": 7218 }, { "epoch": 0.9183310011448925, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.177448272705078, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.877007782459259, "num_tokens": 275357410.0, "step": 7219 }, { "epoch": 0.918458211423483, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.186509132385254, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8749127388000488, "num_tokens": 275388720.0, "step": 7220 }, { "epoch": 0.9185854217020736, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.189233779907227, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8692877888679504, "num_tokens": 275427699.0, "step": 7221 }, { "epoch": 0.918712631980664, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.118369102478027, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8653662204742432, "num_tokens": 275472220.0, "step": 7222 }, { "epoch": 0.9188398422592545, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.226436614990234, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8511616587638855, "num_tokens": 275506693.0, "step": 7223 }, { "epoch": 0.918967052537845, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.21751594543457, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8664369583129883, "num_tokens": 275547295.0, "step": 7224 }, { "epoch": 0.9190942628164356, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.167309761047363, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8590125441551208, "num_tokens": 275591552.0, "step": 7225 }, { "epoch": 0.9192214730950261, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.161298751831055, "learning_rate": 1e-06, "loss": 0.5073, "mean_token_accuracy": 0.8380202054977417, "num_tokens": 275632197.0, "step": 7226 }, { "epoch": 0.9193486833736166, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.106070518493652, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8560841083526611, "num_tokens": 275670680.0, "step": 7227 }, { "epoch": 0.919475893652207, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.203957557678223, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8615274429321289, "num_tokens": 275709315.0, "step": 7228 }, { "epoch": 0.9196031039307976, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.182230949401855, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8626260757446289, "num_tokens": 275740405.0, "step": 7229 }, { "epoch": 0.9197303142093881, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.184050559997559, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8798882961273193, "num_tokens": 275780915.0, "step": 7230 }, { "epoch": 0.9198575244879786, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.141560554504395, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8575274348258972, "num_tokens": 275817832.0, "step": 7231 }, { "epoch": 0.9199847347665692, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.166778564453125, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8670150637626648, "num_tokens": 275857621.0, "step": 7232 }, { "epoch": 0.9201119450451597, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.17263126373291, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8589589595794678, "num_tokens": 275898917.0, "step": 7233 }, { "epoch": 0.9202391553237501, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.178566932678223, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8671348690986633, "num_tokens": 275930151.0, "step": 7234 }, { "epoch": 0.9203663656023406, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.184036254882812, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8698170185089111, "num_tokens": 275966901.0, "step": 7235 }, { "epoch": 0.9204935758809312, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.171478271484375, "learning_rate": 1e-06, "loss": 0.35, "mean_token_accuracy": 0.8846895694732666, "num_tokens": 276002845.0, "step": 7236 }, { "epoch": 0.9206207861595217, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.175018310546875, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8619417548179626, "num_tokens": 276038476.0, "step": 7237 }, { "epoch": 0.9207479964381122, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.188206672668457, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8589768409729004, "num_tokens": 276077867.0, "step": 7238 }, { "epoch": 0.9208752067167028, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.1448974609375, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8521490097045898, "num_tokens": 276123289.0, "step": 7239 }, { "epoch": 0.9210024169952932, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.206518173217773, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8507227301597595, "num_tokens": 276157357.0, "step": 7240 }, { "epoch": 0.9211296272738837, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.151860237121582, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8842151165008545, "num_tokens": 276195548.0, "step": 7241 }, { "epoch": 0.9212568375524742, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.144991874694824, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8651258945465088, "num_tokens": 276233533.0, "step": 7242 }, { "epoch": 0.9213840478310648, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.182367324829102, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8656756281852722, "num_tokens": 276274310.0, "step": 7243 }, { "epoch": 0.9215112581096553, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.181563377380371, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.859663188457489, "num_tokens": 276310272.0, "step": 7244 }, { "epoch": 0.9216384683882458, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.139483451843262, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8569507002830505, "num_tokens": 276350334.0, "step": 7245 }, { "epoch": 0.9217656786668362, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.198344230651855, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8576399087905884, "num_tokens": 276393913.0, "step": 7246 }, { "epoch": 0.9218928889454268, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.206049919128418, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8585600852966309, "num_tokens": 276430368.0, "step": 7247 }, { "epoch": 0.9220200992240173, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.240276336669922, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8562613725662231, "num_tokens": 276470930.0, "step": 7248 }, { "epoch": 0.9221473095026078, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.200892448425293, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8675769567489624, "num_tokens": 276509359.0, "step": 7249 }, { "epoch": 0.9222745197811983, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.182221412658691, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8548287749290466, "num_tokens": 276546980.0, "step": 7250 }, { "epoch": 0.9224017300597889, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.22413444519043, "learning_rate": 1e-06, "loss": 0.4749, "mean_token_accuracy": 0.8506988286972046, "num_tokens": 276586212.0, "step": 7251 }, { "epoch": 0.9225289403383793, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.169156074523926, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8641671538352966, "num_tokens": 276623946.0, "step": 7252 }, { "epoch": 0.9226561506169698, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.215550422668457, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8573358058929443, "num_tokens": 276667567.0, "step": 7253 }, { "epoch": 0.9227833608955603, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.194169044494629, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8562394976615906, "num_tokens": 276708479.0, "step": 7254 }, { "epoch": 0.9229105711741509, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.213396072387695, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8573198318481445, "num_tokens": 276744627.0, "step": 7255 }, { "epoch": 0.9230377814527414, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.200316429138184, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8629457950592041, "num_tokens": 276782830.0, "step": 7256 }, { "epoch": 0.9231649917313319, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.177913665771484, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8608380556106567, "num_tokens": 276825745.0, "step": 7257 }, { "epoch": 0.9232922020099223, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 11.220746994018555, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8673005104064941, "num_tokens": 276866743.0, "step": 7258 }, { "epoch": 0.9234194122885129, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.163702011108398, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8585734963417053, "num_tokens": 276903884.0, "step": 7259 }, { "epoch": 0.9235466225671034, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.207562446594238, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8683047294616699, "num_tokens": 276939142.0, "step": 7260 }, { "epoch": 0.9236738328456939, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.187885284423828, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8638192415237427, "num_tokens": 276979041.0, "step": 7261 }, { "epoch": 0.9238010431242845, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.139908790588379, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8743892908096313, "num_tokens": 277015590.0, "step": 7262 }, { "epoch": 0.923928253402875, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.196659088134766, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8550629019737244, "num_tokens": 277055745.0, "step": 7263 }, { "epoch": 0.9240554636814655, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.173500061035156, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8579635620117188, "num_tokens": 277098856.0, "step": 7264 }, { "epoch": 0.9241826739600559, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.191892623901367, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8732338547706604, "num_tokens": 277139165.0, "step": 7265 }, { "epoch": 0.9243098842386465, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.185782432556152, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8665848970413208, "num_tokens": 277171879.0, "step": 7266 }, { "epoch": 0.924437094517237, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.173746109008789, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8629332780838013, "num_tokens": 277211719.0, "step": 7267 }, { "epoch": 0.9245643047958275, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.139385223388672, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8537918329238892, "num_tokens": 277252675.0, "step": 7268 }, { "epoch": 0.924691515074418, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.25644588470459, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8504748344421387, "num_tokens": 277287529.0, "step": 7269 }, { "epoch": 0.9248187253530086, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.147469520568848, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8611760139465332, "num_tokens": 277328392.0, "step": 7270 }, { "epoch": 0.924945935631599, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.207657814025879, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8718358278274536, "num_tokens": 277368412.0, "step": 7271 }, { "epoch": 0.9250731459101895, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.223118782043457, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8676357269287109, "num_tokens": 277404373.0, "step": 7272 }, { "epoch": 0.92520035618878, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.213008880615234, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8598423600196838, "num_tokens": 277444988.0, "step": 7273 }, { "epoch": 0.9253275664673706, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.197656631469727, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8714540600776672, "num_tokens": 277487501.0, "step": 7274 }, { "epoch": 0.9254547767459611, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.285186767578125, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8591592311859131, "num_tokens": 277529056.0, "step": 7275 }, { "epoch": 0.9255819870245516, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.247942924499512, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8745372891426086, "num_tokens": 277563022.0, "step": 7276 }, { "epoch": 0.925709197303142, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.211777687072754, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8692338466644287, "num_tokens": 277603879.0, "step": 7277 }, { "epoch": 0.9258364075817326, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.19924259185791, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8586035966873169, "num_tokens": 277643711.0, "step": 7278 }, { "epoch": 0.9259636178603231, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.249674797058105, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8620575666427612, "num_tokens": 277680667.0, "step": 7279 }, { "epoch": 0.9260908281389136, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.258217811584473, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.861229419708252, "num_tokens": 277715548.0, "step": 7280 }, { "epoch": 0.9262180384175042, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.250947952270508, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8615468740463257, "num_tokens": 277745287.0, "step": 7281 }, { "epoch": 0.9263452486960947, "ewc_loss": 0.02294921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.215092658996582, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8558059930801392, "num_tokens": 277781484.0, "step": 7282 }, { "epoch": 0.9264724589746851, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.208544731140137, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8650614619255066, "num_tokens": 277818429.0, "step": 7283 }, { "epoch": 0.9265996692532756, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.199851989746094, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8686882257461548, "num_tokens": 277853077.0, "step": 7284 }, { "epoch": 0.9267268795318662, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.174518585205078, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8798455595970154, "num_tokens": 277893997.0, "step": 7285 }, { "epoch": 0.9268540898104567, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.196564674377441, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8519189357757568, "num_tokens": 277934680.0, "step": 7286 }, { "epoch": 0.9269813000890472, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.187472343444824, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.855656087398529, "num_tokens": 277980370.0, "step": 7287 }, { "epoch": 0.9271085103676378, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.195747375488281, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8740564584732056, "num_tokens": 278017257.0, "step": 7288 }, { "epoch": 0.9272357206462282, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.219544410705566, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8623109459877014, "num_tokens": 278054835.0, "step": 7289 }, { "epoch": 0.9273629309248187, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.226730346679688, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8604389429092407, "num_tokens": 278089510.0, "step": 7290 }, { "epoch": 0.9274901412034092, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.218246459960938, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8643474578857422, "num_tokens": 278126179.0, "step": 7291 }, { "epoch": 0.9276173514819998, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.22939682006836, "learning_rate": 1e-06, "loss": 0.4994, "mean_token_accuracy": 0.8402419090270996, "num_tokens": 278170837.0, "step": 7292 }, { "epoch": 0.9277445617605903, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.157032012939453, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8673340082168579, "num_tokens": 278207410.0, "step": 7293 }, { "epoch": 0.9278717720391808, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.229557037353516, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8825720548629761, "num_tokens": 278240495.0, "step": 7294 }, { "epoch": 0.9279989823177712, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.240752220153809, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8540382385253906, "num_tokens": 278278449.0, "step": 7295 }, { "epoch": 0.9281261925963618, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.14012622833252, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8721840977668762, "num_tokens": 278321434.0, "step": 7296 }, { "epoch": 0.9282534028749523, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.162233352661133, "learning_rate": 1e-06, "loss": 0.4623, "mean_token_accuracy": 0.8578781485557556, "num_tokens": 278363730.0, "step": 7297 }, { "epoch": 0.9283806131535428, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.224993705749512, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8641064763069153, "num_tokens": 278395493.0, "step": 7298 }, { "epoch": 0.9285078234321333, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.173993110656738, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8646237850189209, "num_tokens": 278433024.0, "step": 7299 }, { "epoch": 0.9286350337107239, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.212316513061523, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8644225597381592, "num_tokens": 278471311.0, "step": 7300 }, { "epoch": 0.9287622439893143, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.164407730102539, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8825535178184509, "num_tokens": 278509151.0, "step": 7301 }, { "epoch": 0.9288894542679048, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.155712127685547, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8702862858772278, "num_tokens": 278548743.0, "step": 7302 }, { "epoch": 0.9290166645464953, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.186193466186523, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8738821148872375, "num_tokens": 278590722.0, "step": 7303 }, { "epoch": 0.9291438748250859, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.238324165344238, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8504440784454346, "num_tokens": 278628150.0, "step": 7304 }, { "epoch": 0.9292710851036764, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.160920143127441, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.8473550081253052, "num_tokens": 278672543.0, "step": 7305 }, { "epoch": 0.9293982953822669, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 11.229259490966797, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8714277148246765, "num_tokens": 278707234.0, "step": 7306 }, { "epoch": 0.9295255056608573, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.238618850708008, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8651663064956665, "num_tokens": 278734678.0, "step": 7307 }, { "epoch": 0.9296527159394479, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.112931251525879, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.872877836227417, "num_tokens": 278778284.0, "step": 7308 }, { "epoch": 0.9297799262180384, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.218549728393555, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8627045154571533, "num_tokens": 278815363.0, "step": 7309 }, { "epoch": 0.9299071364966289, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.190092086791992, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8509954214096069, "num_tokens": 278860777.0, "step": 7310 }, { "epoch": 0.9300343467752195, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.172053337097168, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8782458305358887, "num_tokens": 278890552.0, "step": 7311 }, { "epoch": 0.93016155705381, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.226261138916016, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8522332906723022, "num_tokens": 278924741.0, "step": 7312 }, { "epoch": 0.9302887673324005, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.197030067443848, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.872020423412323, "num_tokens": 278965185.0, "step": 7313 }, { "epoch": 0.9304159776109909, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.185280799865723, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8525683879852295, "num_tokens": 278998961.0, "step": 7314 }, { "epoch": 0.9305431878895815, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.253361701965332, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8499284982681274, "num_tokens": 279032689.0, "step": 7315 }, { "epoch": 0.930670398168172, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.176126480102539, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.849335789680481, "num_tokens": 279080305.0, "step": 7316 }, { "epoch": 0.9307976084467625, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.186226844787598, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8646914958953857, "num_tokens": 279120581.0, "step": 7317 }, { "epoch": 0.930924818725353, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.225517272949219, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8750219345092773, "num_tokens": 279165050.0, "step": 7318 }, { "epoch": 0.9310520290039436, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.179564476013184, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8670287132263184, "num_tokens": 279209393.0, "step": 7319 }, { "epoch": 0.931179239282534, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.214217185974121, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8507269620895386, "num_tokens": 279250569.0, "step": 7320 }, { "epoch": 0.9313064495611245, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.203764915466309, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8691496253013611, "num_tokens": 279291154.0, "step": 7321 }, { "epoch": 0.931433659839715, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.222347259521484, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8623223304748535, "num_tokens": 279326967.0, "step": 7322 }, { "epoch": 0.9315608701183056, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.167902946472168, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8751243352890015, "num_tokens": 279367129.0, "step": 7323 }, { "epoch": 0.9316880803968961, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.253628730773926, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8692595362663269, "num_tokens": 279403651.0, "step": 7324 }, { "epoch": 0.9318152906754866, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.212638854980469, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8558839559555054, "num_tokens": 279439166.0, "step": 7325 }, { "epoch": 0.931942500954077, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.239033699035645, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8622226119041443, "num_tokens": 279473489.0, "step": 7326 }, { "epoch": 0.9320697112326676, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.270110130310059, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.866152286529541, "num_tokens": 279511709.0, "step": 7327 }, { "epoch": 0.9321969215112581, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.215909004211426, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.851102352142334, "num_tokens": 279551638.0, "step": 7328 }, { "epoch": 0.9323241317898486, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.222563743591309, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.874668538570404, "num_tokens": 279587304.0, "step": 7329 }, { "epoch": 0.9324513420684392, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.190694808959961, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8542377352714539, "num_tokens": 279628138.0, "step": 7330 }, { "epoch": 0.9325785523470297, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.22485065460205, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8686497807502747, "num_tokens": 279663064.0, "step": 7331 }, { "epoch": 0.9327057626256201, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.269733428955078, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8551260232925415, "num_tokens": 279704937.0, "step": 7332 }, { "epoch": 0.9328329729042106, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.232025146484375, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8655960559844971, "num_tokens": 279738215.0, "step": 7333 }, { "epoch": 0.9329601831828012, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.193910598754883, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8461695909500122, "num_tokens": 279771996.0, "step": 7334 }, { "epoch": 0.9330873934613917, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.183860778808594, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8774278163909912, "num_tokens": 279810037.0, "step": 7335 }, { "epoch": 0.9332146037399822, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.230061531066895, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.857450008392334, "num_tokens": 279847394.0, "step": 7336 }, { "epoch": 0.9333418140185727, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.212048530578613, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8497587442398071, "num_tokens": 279879057.0, "step": 7337 }, { "epoch": 0.9334690242971632, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.19616985321045, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8570533990859985, "num_tokens": 279911775.0, "step": 7338 }, { "epoch": 0.9335962345757537, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.247322082519531, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8658856153488159, "num_tokens": 279947661.0, "step": 7339 }, { "epoch": 0.9337234448543442, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.231797218322754, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8645907640457153, "num_tokens": 279977507.0, "step": 7340 }, { "epoch": 0.9338506551329347, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.225258827209473, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8599400520324707, "num_tokens": 280016749.0, "step": 7341 }, { "epoch": 0.9339778654115253, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.237210273742676, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8650646209716797, "num_tokens": 280056092.0, "step": 7342 }, { "epoch": 0.9341050756901158, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.242344856262207, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8756038546562195, "num_tokens": 280089803.0, "step": 7343 }, { "epoch": 0.9342322859687062, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.26346492767334, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8515899181365967, "num_tokens": 280126190.0, "step": 7344 }, { "epoch": 0.9343594962472968, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.214273452758789, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8745046854019165, "num_tokens": 280162743.0, "step": 7345 }, { "epoch": 0.9344867065258873, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.162467002868652, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8628975749015808, "num_tokens": 280203676.0, "step": 7346 }, { "epoch": 0.9346139168044778, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.205768585205078, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8793632388114929, "num_tokens": 280241765.0, "step": 7347 }, { "epoch": 0.9347411270830683, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.22055435180664, "learning_rate": 1e-06, "loss": 0.5227, "mean_token_accuracy": 0.8424211740493774, "num_tokens": 280280103.0, "step": 7348 }, { "epoch": 0.9348683373616589, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.23275089263916, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8469838500022888, "num_tokens": 280314674.0, "step": 7349 }, { "epoch": 0.9349955476402493, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.215441703796387, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8645914793014526, "num_tokens": 280353890.0, "step": 7350 }, { "epoch": 0.9351227579188398, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.24701976776123, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8555660247802734, "num_tokens": 280396403.0, "step": 7351 }, { "epoch": 0.9352499681974303, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.259913444519043, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8632774949073792, "num_tokens": 280436154.0, "step": 7352 }, { "epoch": 0.9353771784760209, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.282910346984863, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8703094720840454, "num_tokens": 280476480.0, "step": 7353 }, { "epoch": 0.9355043887546114, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.248476028442383, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8673549890518188, "num_tokens": 280505976.0, "step": 7354 }, { "epoch": 0.9356315990332019, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.245889663696289, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8596232533454895, "num_tokens": 280548365.0, "step": 7355 }, { "epoch": 0.9357588093117923, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.268735885620117, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8803840279579163, "num_tokens": 280588074.0, "step": 7356 }, { "epoch": 0.9358860195903829, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.286996841430664, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8719464540481567, "num_tokens": 280626915.0, "step": 7357 }, { "epoch": 0.9360132298689734, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.286128044128418, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8579740524291992, "num_tokens": 280660088.0, "step": 7358 }, { "epoch": 0.9361404401475639, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.211904525756836, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8616588115692139, "num_tokens": 280701326.0, "step": 7359 }, { "epoch": 0.9362676504261545, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.205923080444336, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8730096220970154, "num_tokens": 280740256.0, "step": 7360 }, { "epoch": 0.936394860704745, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.267838478088379, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8770729303359985, "num_tokens": 280773515.0, "step": 7361 }, { "epoch": 0.9365220709833355, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.206864356994629, "learning_rate": 1e-06, "loss": 0.4998, "mean_token_accuracy": 0.8385161757469177, "num_tokens": 280817207.0, "step": 7362 }, { "epoch": 0.9366492812619259, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.232522964477539, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8505841493606567, "num_tokens": 280860253.0, "step": 7363 }, { "epoch": 0.9367764915405165, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.264728546142578, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8596631288528442, "num_tokens": 280892634.0, "step": 7364 }, { "epoch": 0.936903701819107, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.271368980407715, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8720433712005615, "num_tokens": 280930912.0, "step": 7365 }, { "epoch": 0.9370309120976975, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.192720413208008, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8673287630081177, "num_tokens": 280971588.0, "step": 7366 }, { "epoch": 0.937158122376288, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.276066780090332, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8706018924713135, "num_tokens": 281009898.0, "step": 7367 }, { "epoch": 0.9372853326548786, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.228897094726562, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8559025526046753, "num_tokens": 281046726.0, "step": 7368 }, { "epoch": 0.937412542933469, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.170160293579102, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.87721848487854, "num_tokens": 281085163.0, "step": 7369 }, { "epoch": 0.9375397532120595, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.218138694763184, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8688774108886719, "num_tokens": 281128660.0, "step": 7370 }, { "epoch": 0.93766696349065, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.245465278625488, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8455027341842651, "num_tokens": 281163780.0, "step": 7371 }, { "epoch": 0.9377941737692406, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.17825698852539, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8711249828338623, "num_tokens": 281204247.0, "step": 7372 }, { "epoch": 0.9379213840478311, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.23305606842041, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8872956037521362, "num_tokens": 281239518.0, "step": 7373 }, { "epoch": 0.9380485943264216, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.224287986755371, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8718456029891968, "num_tokens": 281275916.0, "step": 7374 }, { "epoch": 0.938175804605012, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.229910850524902, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8743425607681274, "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": 11.218210220336914, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8680803775787354, "num_tokens": 281342677.0, "step": 7376 }, { "epoch": 0.9384302251621931, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.246153831481934, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.859159529209137, "num_tokens": 281379545.0, "step": 7377 }, { "epoch": 0.9385574354407836, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.216828346252441, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8454291820526123, "num_tokens": 281424169.0, "step": 7378 }, { "epoch": 0.9386846457193742, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.187298774719238, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8837808966636658, "num_tokens": 281461306.0, "step": 7379 }, { "epoch": 0.9388118559979647, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.209149360656738, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8774100542068481, "num_tokens": 281498753.0, "step": 7380 }, { "epoch": 0.9389390662765551, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.220257759094238, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8725600838661194, "num_tokens": 281537163.0, "step": 7381 }, { "epoch": 0.9390662765551456, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.233165740966797, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8542430996894836, "num_tokens": 281577926.0, "step": 7382 }, { "epoch": 0.9391934868337362, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.225595474243164, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8829445242881775, "num_tokens": 281621283.0, "step": 7383 }, { "epoch": 0.9393206971123267, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.242609977722168, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8668574094772339, "num_tokens": 281658336.0, "step": 7384 }, { "epoch": 0.9394479073909172, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.222975730895996, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8667429089546204, "num_tokens": 281695137.0, "step": 7385 }, { "epoch": 0.9395751176695077, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.209006309509277, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8788918852806091, "num_tokens": 281733570.0, "step": 7386 }, { "epoch": 0.9397023279480982, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.261299133300781, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8657265901565552, "num_tokens": 281771163.0, "step": 7387 }, { "epoch": 0.9398295382266887, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.178609848022461, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8666552305221558, "num_tokens": 281813715.0, "step": 7388 }, { "epoch": 0.9399567485052792, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.24484920501709, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8622162342071533, "num_tokens": 281856921.0, "step": 7389 }, { "epoch": 0.9400839587838697, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.261259078979492, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8559479713439941, "num_tokens": 281893273.0, "step": 7390 }, { "epoch": 0.9402111690624603, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.307973861694336, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8681449890136719, "num_tokens": 281927452.0, "step": 7391 }, { "epoch": 0.9403383793410508, "ewc_loss": 0.023193359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.205265998840332, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.8886522650718689, "num_tokens": 281965359.0, "step": 7392 }, { "epoch": 0.9404655896196412, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.230867385864258, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.86165452003479, "num_tokens": 281999209.0, "step": 7393 }, { "epoch": 0.9405927998982317, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.170516014099121, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8531838059425354, "num_tokens": 282042129.0, "step": 7394 }, { "epoch": 0.9407200101768223, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.268792152404785, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8546364307403564, "num_tokens": 282087941.0, "step": 7395 }, { "epoch": 0.9408472204554128, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.221417427062988, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8692919611930847, "num_tokens": 282123130.0, "step": 7396 }, { "epoch": 0.9409744307340033, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.230350494384766, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8601731061935425, "num_tokens": 282161142.0, "step": 7397 }, { "epoch": 0.9411016410125939, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.210593223571777, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8872908353805542, "num_tokens": 282196363.0, "step": 7398 }, { "epoch": 0.9412288512911843, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.250771522521973, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8459606766700745, "num_tokens": 282233202.0, "step": 7399 }, { "epoch": 0.9413560615697748, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.191326141357422, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.865074634552002, "num_tokens": 282270976.0, "step": 7400 }, { "epoch": 0.9414832718483653, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.226672172546387, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8742682337760925, "num_tokens": 282312223.0, "step": 7401 }, { "epoch": 0.9416104821269559, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.200697898864746, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8481171131134033, "num_tokens": 282350418.0, "step": 7402 }, { "epoch": 0.9417376924055464, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.273027420043945, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8650368452072144, "num_tokens": 282380751.0, "step": 7403 }, { "epoch": 0.9418649026841369, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.213006019592285, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8766723871231079, "num_tokens": 282420248.0, "step": 7404 }, { "epoch": 0.9419921129627273, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.225193977355957, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8683472871780396, "num_tokens": 282457951.0, "step": 7405 }, { "epoch": 0.9421193232413179, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.23551082611084, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8633885979652405, "num_tokens": 282507379.0, "step": 7406 }, { "epoch": 0.9422465335199084, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.212937355041504, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8522437810897827, "num_tokens": 282546905.0, "step": 7407 }, { "epoch": 0.9423737437984989, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.207052230834961, "learning_rate": 1e-06, "loss": 0.5172, "mean_token_accuracy": 0.8357886075973511, "num_tokens": 282591368.0, "step": 7408 }, { "epoch": 0.9425009540770894, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.287297248840332, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8618159294128418, "num_tokens": 282617331.0, "step": 7409 }, { "epoch": 0.94262816435568, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.202139854431152, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8779054880142212, "num_tokens": 282652346.0, "step": 7410 }, { "epoch": 0.9427553746342705, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.197368621826172, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8680782318115234, "num_tokens": 282691564.0, "step": 7411 }, { "epoch": 0.9428825849128609, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.27064037322998, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8743938207626343, "num_tokens": 282728186.0, "step": 7412 }, { "epoch": 0.9430097951914514, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.236083030700684, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8572033643722534, "num_tokens": 282770491.0, "step": 7413 }, { "epoch": 0.943137005470042, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.22628116607666, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8582823872566223, "num_tokens": 282809824.0, "step": 7414 }, { "epoch": 0.9432642157486325, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.295356750488281, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8683032989501953, "num_tokens": 282846735.0, "step": 7415 }, { "epoch": 0.943391426027223, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.190509796142578, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8831944465637207, "num_tokens": 282882242.0, "step": 7416 }, { "epoch": 0.9435186363058136, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.293846130371094, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8514449000358582, "num_tokens": 282918774.0, "step": 7417 }, { "epoch": 0.943645846584404, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.258894920349121, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8563593029975891, "num_tokens": 282957766.0, "step": 7418 }, { "epoch": 0.9437730568629945, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.258621215820312, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8520591855049133, "num_tokens": 282995980.0, "step": 7419 }, { "epoch": 0.943900267141585, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.30264663696289, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8653404712677002, "num_tokens": 283033148.0, "step": 7420 }, { "epoch": 0.9440274774201756, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.23302173614502, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8693732619285583, "num_tokens": 283069781.0, "step": 7421 }, { "epoch": 0.9441546876987661, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.26167106628418, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8562637567520142, "num_tokens": 283107270.0, "step": 7422 }, { "epoch": 0.9442818979773566, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.264178276062012, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8766069412231445, "num_tokens": 283143571.0, "step": 7423 }, { "epoch": 0.944409108255947, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.2232084274292, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8680145740509033, "num_tokens": 283179995.0, "step": 7424 }, { "epoch": 0.9445363185345376, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.229432106018066, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8641825914382935, "num_tokens": 283218370.0, "step": 7425 }, { "epoch": 0.9446635288131281, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.321310997009277, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8581779599189758, "num_tokens": 283247879.0, "step": 7426 }, { "epoch": 0.9447907390917186, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.255146026611328, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8793650269508362, "num_tokens": 283285027.0, "step": 7427 }, { "epoch": 0.9449179493703092, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.189433097839355, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8675131797790527, "num_tokens": 283324934.0, "step": 7428 }, { "epoch": 0.9450451596488997, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.238821029663086, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8718675971031189, "num_tokens": 283357540.0, "step": 7429 }, { "epoch": 0.9451723699274901, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.25129222869873, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8618810176849365, "num_tokens": 283393470.0, "step": 7430 }, { "epoch": 0.9452995802060806, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.290852546691895, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8643026351928711, "num_tokens": 283429916.0, "step": 7431 }, { "epoch": 0.9454267904846712, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.213508605957031, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8533073663711548, "num_tokens": 283467997.0, "step": 7432 }, { "epoch": 0.9455540007632617, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.33037281036377, "learning_rate": 1e-06, "loss": 0.5233, "mean_token_accuracy": 0.8304168581962585, "num_tokens": 283504741.0, "step": 7433 }, { "epoch": 0.9456812110418522, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.252241134643555, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8738168478012085, "num_tokens": 283534535.0, "step": 7434 }, { "epoch": 0.9458084213204427, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.251766204833984, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8757264614105225, "num_tokens": 283571600.0, "step": 7435 }, { "epoch": 0.9459356315990332, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.24558162689209, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8499138355255127, "num_tokens": 283605662.0, "step": 7436 }, { "epoch": 0.9460628418776237, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.260540962219238, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8643791675567627, "num_tokens": 283642868.0, "step": 7437 }, { "epoch": 0.9461900521562142, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.235311508178711, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8552908897399902, "num_tokens": 283675604.0, "step": 7438 }, { "epoch": 0.9463172624348047, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.22933292388916, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8844057321548462, "num_tokens": 283709459.0, "step": 7439 }, { "epoch": 0.9464444727133953, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.219673156738281, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8633188605308533, "num_tokens": 283748020.0, "step": 7440 }, { "epoch": 0.9465716829919858, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.290828704833984, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8588037490844727, "num_tokens": 283778514.0, "step": 7441 }, { "epoch": 0.9466988932705762, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.190835952758789, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8681403994560242, "num_tokens": 283819764.0, "step": 7442 }, { "epoch": 0.9468261035491667, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.294133186340332, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8492199182510376, "num_tokens": 283853897.0, "step": 7443 }, { "epoch": 0.9469533138277573, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.277223587036133, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8775691986083984, "num_tokens": 283891826.0, "step": 7444 }, { "epoch": 0.9470805241063478, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.249358177185059, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8617720007896423, "num_tokens": 283931690.0, "step": 7445 }, { "epoch": 0.9472077343849383, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.27379322052002, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8650146126747131, "num_tokens": 283971107.0, "step": 7446 }, { "epoch": 0.9473349446635289, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.329866409301758, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8714243173599243, "num_tokens": 284005286.0, "step": 7447 }, { "epoch": 0.9474621549421193, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.233673095703125, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8794045448303223, "num_tokens": 284044089.0, "step": 7448 }, { "epoch": 0.9475893652207098, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.264771461486816, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8594273328781128, "num_tokens": 284081172.0, "step": 7449 }, { "epoch": 0.9477165754993003, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.274584770202637, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8714093565940857, "num_tokens": 284113362.0, "step": 7450 }, { "epoch": 0.9478437857778909, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.27806568145752, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8731158971786499, "num_tokens": 284155272.0, "step": 7451 }, { "epoch": 0.9479709960564814, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.29376220703125, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8721351623535156, "num_tokens": 284188218.0, "step": 7452 }, { "epoch": 0.9480982063350719, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.248146057128906, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8683078289031982, "num_tokens": 284228838.0, "step": 7453 }, { "epoch": 0.9482254166136623, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.235380172729492, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8615188598632812, "num_tokens": 284265570.0, "step": 7454 }, { "epoch": 0.9483526268922529, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.266697883605957, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8546069264411926, "num_tokens": 284307062.0, "step": 7455 }, { "epoch": 0.9484798371708434, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.281835556030273, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8619781732559204, "num_tokens": 284340488.0, "step": 7456 }, { "epoch": 0.9486070474494339, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.296875953674316, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8632848262786865, "num_tokens": 284382399.0, "step": 7457 }, { "epoch": 0.9487342577280244, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.254698753356934, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8511521220207214, "num_tokens": 284416391.0, "step": 7458 }, { "epoch": 0.948861468006615, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.264362335205078, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8636869192123413, "num_tokens": 284457049.0, "step": 7459 }, { "epoch": 0.9489886782852054, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.274409294128418, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8622782230377197, "num_tokens": 284494426.0, "step": 7460 }, { "epoch": 0.9491158885637959, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.23261833190918, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8831907510757446, "num_tokens": 284533349.0, "step": 7461 }, { "epoch": 0.9492430988423864, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.270883560180664, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8717626929283142, "num_tokens": 284567777.0, "step": 7462 }, { "epoch": 0.949370309120977, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.313232421875, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8575271964073181, "num_tokens": 284609120.0, "step": 7463 }, { "epoch": 0.9494975193995675, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.247030258178711, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.857088029384613, "num_tokens": 284649221.0, "step": 7464 }, { "epoch": 0.949624729678158, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.290657997131348, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8693574666976929, "num_tokens": 284682595.0, "step": 7465 }, { "epoch": 0.9497519399567486, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.231621742248535, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8602060079574585, "num_tokens": 284722903.0, "step": 7466 }, { "epoch": 0.949879150235339, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.246088027954102, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8701865673065186, "num_tokens": 284759166.0, "step": 7467 }, { "epoch": 0.9500063605139295, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.265913963317871, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8604462146759033, "num_tokens": 284803192.0, "step": 7468 }, { "epoch": 0.95013357079252, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.25406551361084, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8661708235740662, "num_tokens": 284837088.0, "step": 7469 }, { "epoch": 0.9502607810711106, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.281972885131836, "learning_rate": 1e-06, "loss": 0.4948, "mean_token_accuracy": 0.8455792665481567, "num_tokens": 284879641.0, "step": 7470 }, { "epoch": 0.9503879913497011, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.249810218811035, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8528005480766296, "num_tokens": 284917519.0, "step": 7471 }, { "epoch": 0.9505152016282916, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.286602973937988, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8549496531486511, "num_tokens": 284950422.0, "step": 7472 }, { "epoch": 0.950642411906882, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.211132049560547, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8715968132019043, "num_tokens": 284986392.0, "step": 7473 }, { "epoch": 0.9507696221854726, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.247476577758789, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.865537166595459, "num_tokens": 285027684.0, "step": 7474 }, { "epoch": 0.9508968324640631, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.275619506835938, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8559675216674805, "num_tokens": 285059551.0, "step": 7475 }, { "epoch": 0.9510240427426536, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.214699745178223, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8651300668716431, "num_tokens": 285100197.0, "step": 7476 }, { "epoch": 0.9511512530212441, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.29119873046875, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8648113012313843, "num_tokens": 285130115.0, "step": 7477 }, { "epoch": 0.9512784632998347, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.214834213256836, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8648816347122192, "num_tokens": 285169069.0, "step": 7478 }, { "epoch": 0.9514056735784251, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.23754596710205, "learning_rate": 1e-06, "loss": 0.4998, "mean_token_accuracy": 0.8441007733345032, "num_tokens": 285209971.0, "step": 7479 }, { "epoch": 0.9515328838570156, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.211071968078613, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8589060306549072, "num_tokens": 285250024.0, "step": 7480 }, { "epoch": 0.9516600941356061, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.252717971801758, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8607801198959351, "num_tokens": 285284900.0, "step": 7481 }, { "epoch": 0.9517873044141967, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.264827728271484, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8683857917785645, "num_tokens": 285325657.0, "step": 7482 }, { "epoch": 0.9519145146927872, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.238595008850098, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8640676736831665, "num_tokens": 285356362.0, "step": 7483 }, { "epoch": 0.9520417249713777, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.277341842651367, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8635430335998535, "num_tokens": 285394321.0, "step": 7484 }, { "epoch": 0.9521689352499682, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.205683708190918, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8832057118415833, "num_tokens": 285435260.0, "step": 7485 }, { "epoch": 0.9522961455285587, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.269548416137695, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8721741437911987, "num_tokens": 285479449.0, "step": 7486 }, { "epoch": 0.9524233558071492, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.210465431213379, "learning_rate": 1e-06, "loss": 0.4973, "mean_token_accuracy": 0.8443689346313477, "num_tokens": 285520492.0, "step": 7487 }, { "epoch": 0.9525505660857397, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.24643325805664, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.860517144203186, "num_tokens": 285558410.0, "step": 7488 }, { "epoch": 0.9526777763643303, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.259455680847168, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8443186283111572, "num_tokens": 285598871.0, "step": 7489 }, { "epoch": 0.9528049866429208, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.261990547180176, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8572486042976379, "num_tokens": 285641374.0, "step": 7490 }, { "epoch": 0.9529321969215112, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.26017951965332, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8761910200119019, "num_tokens": 285675416.0, "step": 7491 }, { "epoch": 0.9530594072001017, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.266897201538086, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8706040382385254, "num_tokens": 285707491.0, "step": 7492 }, { "epoch": 0.9531866174786923, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.279949188232422, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8528175950050354, "num_tokens": 285747571.0, "step": 7493 }, { "epoch": 0.9533138277572828, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.264155387878418, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.886054277420044, "num_tokens": 285786362.0, "step": 7494 }, { "epoch": 0.9534410380358733, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.287318229675293, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8716925382614136, "num_tokens": 285829401.0, "step": 7495 }, { "epoch": 0.9535682483144639, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.258676528930664, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8618245720863342, "num_tokens": 285867853.0, "step": 7496 }, { "epoch": 0.9536954585930543, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.187618255615234, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8749533891677856, "num_tokens": 285908497.0, "step": 7497 }, { "epoch": 0.9538226688716448, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.29058837890625, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8508909940719604, "num_tokens": 285940528.0, "step": 7498 }, { "epoch": 0.9539498791502353, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.276906967163086, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.864236056804657, "num_tokens": 285980206.0, "step": 7499 }, { "epoch": 0.9540770894288259, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.219686508178711, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8582548499107361, "num_tokens": 286019409.0, "step": 7500 }, { "epoch": 0.9542042997074164, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.296402931213379, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8732254505157471, "num_tokens": 286050955.0, "step": 7501 }, { "epoch": 0.9543315099860069, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.33476734161377, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8686723709106445, "num_tokens": 286087567.0, "step": 7502 }, { "epoch": 0.9544587202645973, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.257051467895508, "learning_rate": 1e-06, "loss": 0.4892, "mean_token_accuracy": 0.8437713980674744, "num_tokens": 286126893.0, "step": 7503 }, { "epoch": 0.9545859305431879, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.27325439453125, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8578103184700012, "num_tokens": 286165818.0, "step": 7504 }, { "epoch": 0.9547131408217784, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.275882720947266, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8666535019874573, "num_tokens": 286208026.0, "step": 7505 }, { "epoch": 0.9548403511003689, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.293610572814941, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.863720178604126, "num_tokens": 286245468.0, "step": 7506 }, { "epoch": 0.9549675613789594, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.302447319030762, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8681373596191406, "num_tokens": 286283955.0, "step": 7507 }, { "epoch": 0.95509477165755, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.281099319458008, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8696441650390625, "num_tokens": 286326560.0, "step": 7508 }, { "epoch": 0.9552219819361404, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.283621788024902, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8645899295806885, "num_tokens": 286365791.0, "step": 7509 }, { "epoch": 0.9553491922147309, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.325451850891113, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.874798059463501, "num_tokens": 286404132.0, "step": 7510 }, { "epoch": 0.9554764024933214, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.243255615234375, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8538154363632202, "num_tokens": 286444828.0, "step": 7511 }, { "epoch": 0.955603612771912, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.31574535369873, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8689694404602051, "num_tokens": 286481725.0, "step": 7512 }, { "epoch": 0.9557308230505025, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.227808952331543, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8650511503219604, "num_tokens": 286520354.0, "step": 7513 }, { "epoch": 0.955858033329093, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.280533790588379, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8622579574584961, "num_tokens": 286564314.0, "step": 7514 }, { "epoch": 0.9559852436076836, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.253334999084473, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8633042573928833, "num_tokens": 286606955.0, "step": 7515 }, { "epoch": 0.956112453886274, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.296242713928223, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8708652257919312, "num_tokens": 286640091.0, "step": 7516 }, { "epoch": 0.9562396641648645, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.295202255249023, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8642269372940063, "num_tokens": 286679090.0, "step": 7517 }, { "epoch": 0.956366874443455, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.225817680358887, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.868337869644165, "num_tokens": 286720260.0, "step": 7518 }, { "epoch": 0.9564940847220456, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.306550025939941, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8697378635406494, "num_tokens": 286763992.0, "step": 7519 }, { "epoch": 0.9566212950006361, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.285613059997559, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8457164764404297, "num_tokens": 286807068.0, "step": 7520 }, { "epoch": 0.9567485052792266, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.248605728149414, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8710771799087524, "num_tokens": 286853020.0, "step": 7521 }, { "epoch": 0.956875715557817, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.309581756591797, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8547530174255371, "num_tokens": 286896098.0, "step": 7522 }, { "epoch": 0.9570029258364076, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.265613555908203, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8486303091049194, "num_tokens": 286938873.0, "step": 7523 }, { "epoch": 0.9571301361149981, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.32929801940918, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8687343597412109, "num_tokens": 286972941.0, "step": 7524 }, { "epoch": 0.9572573463935886, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 11.270391464233398, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8539672493934631, "num_tokens": 287010818.0, "step": 7525 }, { "epoch": 0.9573845566721791, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.23938274383545, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8690348267555237, "num_tokens": 287053841.0, "step": 7526 }, { "epoch": 0.9575117669507697, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.259268760681152, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8701577186584473, "num_tokens": 287090900.0, "step": 7527 }, { "epoch": 0.9576389772293601, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.33164119720459, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.862478494644165, "num_tokens": 287131203.0, "step": 7528 }, { "epoch": 0.9577661875079506, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.252960205078125, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8747857809066772, "num_tokens": 287165949.0, "step": 7529 }, { "epoch": 0.9578933977865411, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.321627616882324, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8639959692955017, "num_tokens": 287203376.0, "step": 7530 }, { "epoch": 0.9580206080651317, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.263836860656738, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8547621369361877, "num_tokens": 287240391.0, "step": 7531 }, { "epoch": 0.9581478183437222, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.313926696777344, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8613815307617188, "num_tokens": 287278485.0, "step": 7532 }, { "epoch": 0.9582750286223127, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.284310340881348, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8597863912582397, "num_tokens": 287313626.0, "step": 7533 }, { "epoch": 0.9584022389009031, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.22252368927002, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8798970580101013, "num_tokens": 287344532.0, "step": 7534 }, { "epoch": 0.9585294491794937, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.28867244720459, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8446404933929443, "num_tokens": 287383069.0, "step": 7535 }, { "epoch": 0.9586566594580842, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.254630088806152, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.863523006439209, "num_tokens": 287418554.0, "step": 7536 }, { "epoch": 0.9587838697366747, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.307412147521973, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8625662326812744, "num_tokens": 287452449.0, "step": 7537 }, { "epoch": 0.9589110800152653, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.276054382324219, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8499127626419067, "num_tokens": 287492247.0, "step": 7538 }, { "epoch": 0.9590382902938558, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.2936429977417, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8704044222831726, "num_tokens": 287528651.0, "step": 7539 }, { "epoch": 0.9591655005724462, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.29863452911377, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8538632988929749, "num_tokens": 287565837.0, "step": 7540 }, { "epoch": 0.9592927108510367, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.334078788757324, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8630818724632263, "num_tokens": 287599469.0, "step": 7541 }, { "epoch": 0.9594199211296273, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.283394813537598, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8619810342788696, "num_tokens": 287637104.0, "step": 7542 }, { "epoch": 0.9595471314082178, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.279120445251465, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8645762205123901, "num_tokens": 287679134.0, "step": 7543 }, { "epoch": 0.9596743416868083, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.216994285583496, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8716254830360413, "num_tokens": 287720340.0, "step": 7544 }, { "epoch": 0.9598015519653988, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.322986602783203, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8680358529090881, "num_tokens": 287760931.0, "step": 7545 }, { "epoch": 0.9599287622439893, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.270886421203613, "learning_rate": 1e-06, "loss": 0.4763, "mean_token_accuracy": 0.8510684967041016, "num_tokens": 287801723.0, "step": 7546 }, { "epoch": 0.9600559725225798, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.292548179626465, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8624927997589111, "num_tokens": 287837010.0, "step": 7547 }, { "epoch": 0.9601831828011703, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.275693893432617, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8655929565429688, "num_tokens": 287874381.0, "step": 7548 }, { "epoch": 0.9603103930797608, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.247446060180664, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8567777872085571, "num_tokens": 287916673.0, "step": 7549 }, { "epoch": 0.9604376033583514, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.295586585998535, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8591837286949158, "num_tokens": 287956764.0, "step": 7550 }, { "epoch": 0.9605648136369419, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.2776460647583, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8721468448638916, "num_tokens": 287996373.0, "step": 7551 }, { "epoch": 0.9606920239155323, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.305473327636719, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8593409061431885, "num_tokens": 288033671.0, "step": 7552 }, { "epoch": 0.9608192341941229, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.252889633178711, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8724638223648071, "num_tokens": 288068254.0, "step": 7553 }, { "epoch": 0.9609464444727134, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.3357515335083, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8495973348617554, "num_tokens": 288102833.0, "step": 7554 }, { "epoch": 0.9610736547513039, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.274157524108887, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8768194913864136, "num_tokens": 288135927.0, "step": 7555 }, { "epoch": 0.9612008650298944, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.306794166564941, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8839678764343262, "num_tokens": 288173842.0, "step": 7556 }, { "epoch": 0.961328075308485, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.314291954040527, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8702526092529297, "num_tokens": 288215007.0, "step": 7557 }, { "epoch": 0.9614552855870754, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.287019729614258, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8538687229156494, "num_tokens": 288251598.0, "step": 7558 }, { "epoch": 0.9615824958656659, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.313758850097656, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8649827241897583, "num_tokens": 288292811.0, "step": 7559 }, { "epoch": 0.9617097061442564, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.292485237121582, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8687183856964111, "num_tokens": 288337790.0, "step": 7560 }, { "epoch": 0.961836916422847, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.278329849243164, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8610069751739502, "num_tokens": 288373347.0, "step": 7561 }, { "epoch": 0.9619641267014375, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.303985595703125, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8839501142501831, "num_tokens": 288405650.0, "step": 7562 }, { "epoch": 0.962091336980028, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.291675567626953, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8699905872344971, "num_tokens": 288443849.0, "step": 7563 }, { "epoch": 0.9622185472586186, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.281970977783203, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.8439024090766907, "num_tokens": 288487557.0, "step": 7564 }, { "epoch": 0.962345757537209, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.326613426208496, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8553242683410645, "num_tokens": 288529932.0, "step": 7565 }, { "epoch": 0.9624729678157995, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.298083305358887, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8593581914901733, "num_tokens": 288568276.0, "step": 7566 }, { "epoch": 0.96260017809439, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.292808532714844, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.866743803024292, "num_tokens": 288602387.0, "step": 7567 }, { "epoch": 0.9627273883729806, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.31629467010498, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8721427321434021, "num_tokens": 288638157.0, "step": 7568 }, { "epoch": 0.9628545986515711, "ewc_loss": 0.0234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 11.30827522277832, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8607652187347412, "num_tokens": 288675054.0, "step": 7569 }, { "epoch": 0.9629818089301616, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.318205833435059, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8699132800102234, "num_tokens": 288714222.0, "step": 7570 }, { "epoch": 0.963109019208752, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.254745483398438, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8543285131454468, "num_tokens": 288752772.0, "step": 7571 }, { "epoch": 0.9632362294873426, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.281597137451172, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.847164511680603, "num_tokens": 288792067.0, "step": 7572 }, { "epoch": 0.9633634397659331, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.30405330657959, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.876133918762207, "num_tokens": 288826172.0, "step": 7573 }, { "epoch": 0.9634906500445236, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.319046020507812, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8746845126152039, "num_tokens": 288863774.0, "step": 7574 }, { "epoch": 0.9636178603231141, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.28288745880127, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8575718402862549, "num_tokens": 288901892.0, "step": 7575 }, { "epoch": 0.9637450706017047, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.277349472045898, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8531641960144043, "num_tokens": 288941470.0, "step": 7576 }, { "epoch": 0.9638722808802951, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.27366828918457, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8509764671325684, "num_tokens": 288988616.0, "step": 7577 }, { "epoch": 0.9639994911588856, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.329371452331543, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8804383277893066, "num_tokens": 289024589.0, "step": 7578 }, { "epoch": 0.9641267014374761, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.204511642456055, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8674260377883911, "num_tokens": 289064334.0, "step": 7579 }, { "epoch": 0.9642539117160667, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.30673599243164, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8530740737915039, "num_tokens": 289099385.0, "step": 7580 }, { "epoch": 0.9643811219946572, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.277576446533203, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8675265312194824, "num_tokens": 289139143.0, "step": 7581 }, { "epoch": 0.9645083322732477, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.278597831726074, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8588534593582153, "num_tokens": 289179303.0, "step": 7582 }, { "epoch": 0.9646355425518381, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.327634811401367, "learning_rate": 1e-06, "loss": 0.4509, "mean_token_accuracy": 0.8568118810653687, "num_tokens": 289214218.0, "step": 7583 }, { "epoch": 0.9647627528304287, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.303912162780762, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8605260252952576, "num_tokens": 289240991.0, "step": 7584 }, { "epoch": 0.9648899631090192, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.347960472106934, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.857245922088623, "num_tokens": 289273374.0, "step": 7585 }, { "epoch": 0.9650171733876097, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.242100715637207, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8501827716827393, "num_tokens": 289314904.0, "step": 7586 }, { "epoch": 0.9651443836662003, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.282230377197266, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8570110201835632, "num_tokens": 289358110.0, "step": 7587 }, { "epoch": 0.9652715939447908, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.327435493469238, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8543788194656372, "num_tokens": 289391581.0, "step": 7588 }, { "epoch": 0.9653988042233812, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.290914535522461, "learning_rate": 1e-06, "loss": 0.499, "mean_token_accuracy": 0.8390772342681885, "num_tokens": 289432280.0, "step": 7589 }, { "epoch": 0.9655260145019717, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.34325122833252, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.847719669342041, "num_tokens": 289473222.0, "step": 7590 }, { "epoch": 0.9656532247805623, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.294243812561035, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8598513603210449, "num_tokens": 289512898.0, "step": 7591 }, { "epoch": 0.9657804350591528, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.293771743774414, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8689624071121216, "num_tokens": 289554394.0, "step": 7592 }, { "epoch": 0.9659076453377433, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.328412055969238, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8592929840087891, "num_tokens": 289594871.0, "step": 7593 }, { "epoch": 0.9660348556163338, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.349453926086426, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8530238270759583, "num_tokens": 289629165.0, "step": 7594 }, { "epoch": 0.9661620658949243, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.247310638427734, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8628560304641724, "num_tokens": 289670131.0, "step": 7595 }, { "epoch": 0.9662892761735148, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.308076858520508, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8666183352470398, "num_tokens": 289705238.0, "step": 7596 }, { "epoch": 0.9664164864521053, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.297220230102539, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8661673069000244, "num_tokens": 289740146.0, "step": 7597 }, { "epoch": 0.9665436967306958, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.297017097473145, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8681873083114624, "num_tokens": 289781145.0, "step": 7598 }, { "epoch": 0.9666709070092864, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.283117294311523, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8464522361755371, "num_tokens": 289816347.0, "step": 7599 }, { "epoch": 0.9667981172878769, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.320290565490723, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8493186235427856, "num_tokens": 289852974.0, "step": 7600 }, { "epoch": 0.9669253275664673, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.272403717041016, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8520260453224182, "num_tokens": 289894553.0, "step": 7601 }, { "epoch": 0.9670525378450578, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.30065631866455, "learning_rate": 1e-06, "loss": 0.4887, "mean_token_accuracy": 0.8458627462387085, "num_tokens": 289930641.0, "step": 7602 }, { "epoch": 0.9671797481236484, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.260845184326172, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8569196462631226, "num_tokens": 289972690.0, "step": 7603 }, { "epoch": 0.9673069584022389, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.331830978393555, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8577543497085571, "num_tokens": 290006518.0, "step": 7604 }, { "epoch": 0.9674341686808294, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.280487060546875, "learning_rate": 1e-06, "loss": 0.4789, "mean_token_accuracy": 0.8501454591751099, "num_tokens": 290045206.0, "step": 7605 }, { "epoch": 0.96756137895942, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.319211959838867, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8678755164146423, "num_tokens": 290086976.0, "step": 7606 }, { "epoch": 0.9676885892380104, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.282979011535645, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8754855394363403, "num_tokens": 290126894.0, "step": 7607 }, { "epoch": 0.9678157995166009, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.299206733703613, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8764156699180603, "num_tokens": 290161684.0, "step": 7608 }, { "epoch": 0.9679430097951914, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.306617736816406, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8657204508781433, "num_tokens": 290203407.0, "step": 7609 }, { "epoch": 0.968070220073782, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.345477104187012, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8697885274887085, "num_tokens": 290238645.0, "step": 7610 }, { "epoch": 0.9681974303523725, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.281023025512695, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8628859519958496, "num_tokens": 290280279.0, "step": 7611 }, { "epoch": 0.968324640630963, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.34134292602539, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8618621826171875, "num_tokens": 290317739.0, "step": 7612 }, { "epoch": 0.9684518509095535, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.276153564453125, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8615133762359619, "num_tokens": 290360384.0, "step": 7613 }, { "epoch": 0.968579061188144, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.340705871582031, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.870823860168457, "num_tokens": 290398706.0, "step": 7614 }, { "epoch": 0.9687062714667345, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.287331581115723, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8769271373748779, "num_tokens": 290439524.0, "step": 7615 }, { "epoch": 0.968833481745325, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.359418869018555, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8582479953765869, "num_tokens": 290472125.0, "step": 7616 }, { "epoch": 0.9689606920239155, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.288960456848145, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8476577401161194, "num_tokens": 290514348.0, "step": 7617 }, { "epoch": 0.9690879023025061, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.320423126220703, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8577595949172974, "num_tokens": 290551699.0, "step": 7618 }, { "epoch": 0.9692151125810966, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.28385066986084, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8654530644416809, "num_tokens": 290588644.0, "step": 7619 }, { "epoch": 0.969342322859687, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.291810035705566, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8655247688293457, "num_tokens": 290629546.0, "step": 7620 }, { "epoch": 0.9694695331382776, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.251215934753418, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8716397285461426, "num_tokens": 290665527.0, "step": 7621 }, { "epoch": 0.9695967434168681, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.320176124572754, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8552886247634888, "num_tokens": 290706667.0, "step": 7622 }, { "epoch": 0.9697239536954586, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.320890426635742, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8579646348953247, "num_tokens": 290745698.0, "step": 7623 }, { "epoch": 0.9698511639740491, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.298974990844727, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8687306642532349, "num_tokens": 290777983.0, "step": 7624 }, { "epoch": 0.9699783742526397, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.3167724609375, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8570584654808044, "num_tokens": 290814184.0, "step": 7625 }, { "epoch": 0.9701055845312301, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.290401458740234, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8541860580444336, "num_tokens": 290854054.0, "step": 7626 }, { "epoch": 0.9702327948098206, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.29526138305664, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8625460863113403, "num_tokens": 290891325.0, "step": 7627 }, { "epoch": 0.9703600050884111, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.264780044555664, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8621792793273926, "num_tokens": 290930101.0, "step": 7628 }, { "epoch": 0.9704872153670017, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.27856731414795, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8634319305419922, "num_tokens": 290964691.0, "step": 7629 }, { "epoch": 0.9706144256455922, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.275081634521484, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8590927124023438, "num_tokens": 291008140.0, "step": 7630 }, { "epoch": 0.9707416359241827, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.300224304199219, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8567924499511719, "num_tokens": 291041484.0, "step": 7631 }, { "epoch": 0.9708688462027731, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.323272705078125, "learning_rate": 1e-06, "loss": 0.4919, "mean_token_accuracy": 0.8440043926239014, "num_tokens": 291082337.0, "step": 7632 }, { "epoch": 0.9709960564813637, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.278274536132812, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8619977235794067, "num_tokens": 291115921.0, "step": 7633 }, { "epoch": 0.9711232667599542, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.247894287109375, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8706859946250916, "num_tokens": 291154367.0, "step": 7634 }, { "epoch": 0.9712504770385447, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.294976234436035, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8634001612663269, "num_tokens": 291198919.0, "step": 7635 }, { "epoch": 0.9713776873171353, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.315582275390625, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8524168729782104, "num_tokens": 291237062.0, "step": 7636 }, { "epoch": 0.9715048975957258, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.278345108032227, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8746051788330078, "num_tokens": 291275676.0, "step": 7637 }, { "epoch": 0.9716321078743162, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.312751770019531, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8556161522865295, "num_tokens": 291316785.0, "step": 7638 }, { "epoch": 0.9717593181529067, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.271383285522461, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8604148030281067, "num_tokens": 291360448.0, "step": 7639 }, { "epoch": 0.9718865284314973, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.314937591552734, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8605469465255737, "num_tokens": 291401650.0, "step": 7640 }, { "epoch": 0.9720137387100878, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.262580871582031, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8636422157287598, "num_tokens": 291439736.0, "step": 7641 }, { "epoch": 0.9721409489886783, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.23884105682373, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.86793053150177, "num_tokens": 291479077.0, "step": 7642 }, { "epoch": 0.9722681592672688, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.276046752929688, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8734692335128784, "num_tokens": 291512470.0, "step": 7643 }, { "epoch": 0.9723953695458593, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.308377265930176, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.873624324798584, "num_tokens": 291552797.0, "step": 7644 }, { "epoch": 0.9725225798244498, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.256669998168945, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8667169809341431, "num_tokens": 291589885.0, "step": 7645 }, { "epoch": 0.9726497901030403, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.27224063873291, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8620807528495789, "num_tokens": 291633005.0, "step": 7646 }, { "epoch": 0.9727770003816308, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.309627532958984, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8606564998626709, "num_tokens": 291666083.0, "step": 7647 }, { "epoch": 0.9729042106602214, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.30341911315918, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8697273135185242, "num_tokens": 291706421.0, "step": 7648 }, { "epoch": 0.9730314209388119, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.321453094482422, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8688345551490784, "num_tokens": 291744641.0, "step": 7649 }, { "epoch": 0.9731586312174023, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.269095420837402, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8894503116607666, "num_tokens": 291780586.0, "step": 7650 }, { "epoch": 0.9732858414959928, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.299849510192871, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.859744668006897, "num_tokens": 291820482.0, "step": 7651 }, { "epoch": 0.9734130517745834, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.2997407913208, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8722583651542664, "num_tokens": 291858568.0, "step": 7652 }, { "epoch": 0.9735402620531739, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.340324401855469, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8546836972236633, "num_tokens": 291889383.0, "step": 7653 }, { "epoch": 0.9736674723317644, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.302751541137695, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8723524212837219, "num_tokens": 291927541.0, "step": 7654 }, { "epoch": 0.973794682610355, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.303009033203125, "learning_rate": 1e-06, "loss": 0.467, "mean_token_accuracy": 0.8476008176803589, "num_tokens": 291969074.0, "step": 7655 }, { "epoch": 0.9739218928889454, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.36187744140625, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8612962365150452, "num_tokens": 292008311.0, "step": 7656 }, { "epoch": 0.9740491031675359, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.2746000289917, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8552071452140808, "num_tokens": 292051810.0, "step": 7657 }, { "epoch": 0.9741763134461264, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.264471054077148, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8703369498252869, "num_tokens": 292090411.0, "step": 7658 }, { "epoch": 0.974303523724717, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.295004844665527, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8558588027954102, "num_tokens": 292124411.0, "step": 7659 }, { "epoch": 0.9744307340033075, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.294280052185059, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8627493381500244, "num_tokens": 292166428.0, "step": 7660 }, { "epoch": 0.974557944281898, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.284581184387207, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8655043244361877, "num_tokens": 292202687.0, "step": 7661 }, { "epoch": 0.9746851545604885, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.318381309509277, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8729168176651001, "num_tokens": 292246364.0, "step": 7662 }, { "epoch": 0.974812364839079, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.301794052124023, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8640866279602051, "num_tokens": 292285789.0, "step": 7663 }, { "epoch": 0.9749395751176695, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.34467887878418, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8589149713516235, "num_tokens": 292320755.0, "step": 7664 }, { "epoch": 0.97506678539626, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.301058769226074, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8723515868186951, "num_tokens": 292362818.0, "step": 7665 }, { "epoch": 0.9751939956748505, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.335658073425293, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8553422689437866, "num_tokens": 292404016.0, "step": 7666 }, { "epoch": 0.9753212059534411, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.323188781738281, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.8497191071510315, "num_tokens": 292442117.0, "step": 7667 }, { "epoch": 0.9754484162320316, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.339555740356445, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8712091445922852, "num_tokens": 292472523.0, "step": 7668 }, { "epoch": 0.975575626510622, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.278164863586426, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.867760419845581, "num_tokens": 292508571.0, "step": 7669 }, { "epoch": 0.9757028367892125, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.25039291381836, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8595631122589111, "num_tokens": 292545358.0, "step": 7670 }, { "epoch": 0.9758300470678031, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.311389923095703, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8478215932846069, "num_tokens": 292583962.0, "step": 7671 }, { "epoch": 0.9759572573463936, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.293266296386719, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8625885844230652, "num_tokens": 292616186.0, "step": 7672 }, { "epoch": 0.9760844676249841, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.247462272644043, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8586006760597229, "num_tokens": 292657378.0, "step": 7673 }, { "epoch": 0.9762116779035747, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.322746276855469, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8770180940628052, "num_tokens": 292692041.0, "step": 7674 }, { "epoch": 0.9763388881821651, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.353599548339844, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8609793186187744, "num_tokens": 292725062.0, "step": 7675 }, { "epoch": 0.9764660984607556, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.304611206054688, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8710407018661499, "num_tokens": 292766522.0, "step": 7676 }, { "epoch": 0.9765933087393461, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.230916023254395, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8634427785873413, "num_tokens": 292811371.0, "step": 7677 }, { "epoch": 0.9767205190179367, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.331175804138184, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8637068867683411, "num_tokens": 292850237.0, "step": 7678 }, { "epoch": 0.9768477292965272, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.323784828186035, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8665532469749451, "num_tokens": 292887291.0, "step": 7679 }, { "epoch": 0.9769749395751177, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.291773796081543, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.876105010509491, "num_tokens": 292928396.0, "step": 7680 }, { "epoch": 0.9771021498537081, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.350768089294434, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8668699860572815, "num_tokens": 292962748.0, "step": 7681 }, { "epoch": 0.9772293601322987, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.316837310791016, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8750849962234497, "num_tokens": 293002330.0, "step": 7682 }, { "epoch": 0.9773565704108892, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.273858070373535, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8634033203125, "num_tokens": 293038288.0, "step": 7683 }, { "epoch": 0.9774837806894797, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.350417137145996, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8666325807571411, "num_tokens": 293078121.0, "step": 7684 }, { "epoch": 0.9776109909680702, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.278016090393066, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8603110313415527, "num_tokens": 293116846.0, "step": 7685 }, { "epoch": 0.9777382012466608, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.355106353759766, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8591324090957642, "num_tokens": 293162237.0, "step": 7686 }, { "epoch": 0.9778654115252512, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.289436340332031, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8612967729568481, "num_tokens": 293196631.0, "step": 7687 }, { "epoch": 0.9779926218038417, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.2789306640625, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8736599087715149, "num_tokens": 293238883.0, "step": 7688 }, { "epoch": 0.9781198320824323, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.336965560913086, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8538167476654053, "num_tokens": 293270802.0, "step": 7689 }, { "epoch": 0.9782470423610228, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.325121879577637, "learning_rate": 1e-06, "loss": 0.4937, "mean_token_accuracy": 0.8494001626968384, "num_tokens": 293314910.0, "step": 7690 }, { "epoch": 0.9783742526396133, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.248442649841309, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8760262727737427, "num_tokens": 293353083.0, "step": 7691 }, { "epoch": 0.9785014629182038, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.302153587341309, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8661477565765381, "num_tokens": 293389473.0, "step": 7692 }, { "epoch": 0.9786286731967943, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.33103084564209, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8771330118179321, "num_tokens": 293426893.0, "step": 7693 }, { "epoch": 0.9787558834753848, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.265623092651367, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8689392805099487, "num_tokens": 293468116.0, "step": 7694 }, { "epoch": 0.9788830937539753, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.357449531555176, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8664641380310059, "num_tokens": 293503050.0, "step": 7695 }, { "epoch": 0.9790103040325658, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.307208061218262, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.853136420249939, "num_tokens": 293541673.0, "step": 7696 }, { "epoch": 0.9791375143111564, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.317819595336914, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.861046552658081, "num_tokens": 293578472.0, "step": 7697 }, { "epoch": 0.9792647245897469, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.351354598999023, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8664793968200684, "num_tokens": 293611204.0, "step": 7698 }, { "epoch": 0.9793919348683373, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.333464622497559, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8534045815467834, "num_tokens": 293651352.0, "step": 7699 }, { "epoch": 0.9795191451469278, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.286983489990234, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8741872310638428, "num_tokens": 293687396.0, "step": 7700 }, { "epoch": 0.9796463554255184, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.367483139038086, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8594034910202026, "num_tokens": 293725083.0, "step": 7701 }, { "epoch": 0.9797735657041089, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.259020805358887, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8642288446426392, "num_tokens": 293766136.0, "step": 7702 }, { "epoch": 0.9799007759826994, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.301361083984375, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8470960855484009, "num_tokens": 293805810.0, "step": 7703 }, { "epoch": 0.98002798626129, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.34748363494873, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8556450605392456, "num_tokens": 293839041.0, "step": 7704 }, { "epoch": 0.9801551965398804, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.268220901489258, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8783807754516602, "num_tokens": 293876541.0, "step": 7705 }, { "epoch": 0.9802824068184709, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.2982759475708, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8775362968444824, "num_tokens": 293918141.0, "step": 7706 }, { "epoch": 0.9804096170970614, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.330696105957031, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8529713749885559, "num_tokens": 293958078.0, "step": 7707 }, { "epoch": 0.980536827375652, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.338235855102539, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8737285733222961, "num_tokens": 293990347.0, "step": 7708 }, { "epoch": 0.9806640376542425, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.314369201660156, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8690356016159058, "num_tokens": 294032136.0, "step": 7709 }, { "epoch": 0.980791247932833, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.395685195922852, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.851791501045227, "num_tokens": 294065136.0, "step": 7710 }, { "epoch": 0.9809184582114235, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.316609382629395, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8772491216659546, "num_tokens": 294104539.0, "step": 7711 }, { "epoch": 0.981045668490014, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.254171371459961, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8619298338890076, "num_tokens": 294148962.0, "step": 7712 }, { "epoch": 0.9811728787686045, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.36996841430664, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8669698238372803, "num_tokens": 294181866.0, "step": 7713 }, { "epoch": 0.981300089047195, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.308262825012207, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8728469610214233, "num_tokens": 294223558.0, "step": 7714 }, { "epoch": 0.9814272993257855, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.330766677856445, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8683074116706848, "num_tokens": 294260487.0, "step": 7715 }, { "epoch": 0.9815545096043761, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.34843921661377, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8653026223182678, "num_tokens": 294302642.0, "step": 7716 }, { "epoch": 0.9816817198829666, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.362579345703125, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.85498046875, "num_tokens": 294338376.0, "step": 7717 }, { "epoch": 0.981808930161557, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.366436958312988, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8587019443511963, "num_tokens": 294372089.0, "step": 7718 }, { "epoch": 0.9819361404401475, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.346123695373535, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8695491552352905, "num_tokens": 294403801.0, "step": 7719 }, { "epoch": 0.9820633507187381, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.337854385375977, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8668563365936279, "num_tokens": 294440301.0, "step": 7720 }, { "epoch": 0.9821905609973286, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.358161926269531, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8640578985214233, "num_tokens": 294472967.0, "step": 7721 }, { "epoch": 0.9823177712759191, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.331782341003418, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8549121618270874, "num_tokens": 294518248.0, "step": 7722 }, { "epoch": 0.9824449815545097, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.32341194152832, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8619045615196228, "num_tokens": 294553899.0, "step": 7723 }, { "epoch": 0.9825721918331001, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.396702766418457, "learning_rate": 1e-06, "loss": 0.4812, "mean_token_accuracy": 0.8419485092163086, "num_tokens": 294591382.0, "step": 7724 }, { "epoch": 0.9826994021116906, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.324252128601074, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8652026653289795, "num_tokens": 294629606.0, "step": 7725 }, { "epoch": 0.9828266123902811, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.339466094970703, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8664506673812866, "num_tokens": 294669417.0, "step": 7726 }, { "epoch": 0.9829538226688717, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.37668228149414, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8734180927276611, "num_tokens": 294709876.0, "step": 7727 }, { "epoch": 0.9830810329474622, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.348917007446289, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8551338315010071, "num_tokens": 294750565.0, "step": 7728 }, { "epoch": 0.9832082432260527, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.306607246398926, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.864745557308197, "num_tokens": 294790586.0, "step": 7729 }, { "epoch": 0.9833354535046431, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.346440315246582, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8624518513679504, "num_tokens": 294830176.0, "step": 7730 }, { "epoch": 0.9834626637832337, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.36026382446289, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8516570329666138, "num_tokens": 294874199.0, "step": 7731 }, { "epoch": 0.9835898740618242, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.306046485900879, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8712406158447266, "num_tokens": 294914219.0, "step": 7732 }, { "epoch": 0.9837170843404147, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.397727012634277, "learning_rate": 1e-06, "loss": 0.488, "mean_token_accuracy": 0.8416922092437744, "num_tokens": 294950726.0, "step": 7733 }, { "epoch": 0.9838442946190052, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.3290376663208, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8635575771331787, "num_tokens": 294993815.0, "step": 7734 }, { "epoch": 0.9839715048975958, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.32374095916748, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8722228407859802, "num_tokens": 295029275.0, "step": 7735 }, { "epoch": 0.9840987151761862, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.383026123046875, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8629767894744873, "num_tokens": 295066166.0, "step": 7736 }, { "epoch": 0.9842259254547767, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.37578010559082, "learning_rate": 1e-06, "loss": 0.5036, "mean_token_accuracy": 0.8413254022598267, "num_tokens": 295100436.0, "step": 7737 }, { "epoch": 0.9843531357333672, "ewc_loss": 0.023681640625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.306918144226074, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.864775538444519, "num_tokens": 295141756.0, "step": 7738 }, { "epoch": 0.9844803460119578, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.35521411895752, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8809313774108887, "num_tokens": 295178894.0, "step": 7739 }, { "epoch": 0.9846075562905483, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.347545623779297, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8531625270843506, "num_tokens": 295219133.0, "step": 7740 }, { "epoch": 0.9847347665691388, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.32615852355957, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8647102117538452, "num_tokens": 295260322.0, "step": 7741 }, { "epoch": 0.9848619768477292, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.329604148864746, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8579883575439453, "num_tokens": 295299675.0, "step": 7742 }, { "epoch": 0.9849891871263198, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.352495193481445, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8700379729270935, "num_tokens": 295336281.0, "step": 7743 }, { "epoch": 0.9851163974049103, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.260811805725098, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8608417510986328, "num_tokens": 295376932.0, "step": 7744 }, { "epoch": 0.9852436076835008, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.360857009887695, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.859494686126709, "num_tokens": 295421904.0, "step": 7745 }, { "epoch": 0.9853708179620914, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.369769096374512, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.866398811340332, "num_tokens": 295456688.0, "step": 7746 }, { "epoch": 0.9854980282406819, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.289029121398926, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8806363344192505, "num_tokens": 295500022.0, "step": 7747 }, { "epoch": 0.9856252385192723, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.348505020141602, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8839129209518433, "num_tokens": 295538106.0, "step": 7748 }, { "epoch": 0.9857524487978628, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.32397174835205, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8592238426208496, "num_tokens": 295572465.0, "step": 7749 }, { "epoch": 0.9858796590764534, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.340474128723145, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8698452711105347, "num_tokens": 295614577.0, "step": 7750 }, { "epoch": 0.9860068693550439, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.317041397094727, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8683434724807739, "num_tokens": 295654780.0, "step": 7751 }, { "epoch": 0.9861340796336344, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.321232795715332, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8624576926231384, "num_tokens": 295695880.0, "step": 7752 }, { "epoch": 0.986261289912225, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.35501766204834, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8638782501220703, "num_tokens": 295738176.0, "step": 7753 }, { "epoch": 0.9863885001908154, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 11.278759002685547, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8535006642341614, "num_tokens": 295777562.0, "step": 7754 }, { "epoch": 0.9865157104694059, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.346457481384277, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8809012174606323, "num_tokens": 295815013.0, "step": 7755 }, { "epoch": 0.9866429207479964, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.30446720123291, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8526115417480469, "num_tokens": 295857751.0, "step": 7756 }, { "epoch": 0.986770131026587, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.36454963684082, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8692048788070679, "num_tokens": 295896033.0, "step": 7757 }, { "epoch": 0.9868973413051775, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.334000587463379, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8662288784980774, "num_tokens": 295936838.0, "step": 7758 }, { "epoch": 0.987024551583768, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.317092895507812, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8694864511489868, "num_tokens": 295972307.0, "step": 7759 }, { "epoch": 0.9871517618623584, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.40073013305664, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8617082834243774, "num_tokens": 296009966.0, "step": 7760 }, { "epoch": 0.987278972140949, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.352350234985352, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8609798550605774, "num_tokens": 296047874.0, "step": 7761 }, { "epoch": 0.9874061824195395, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.323891639709473, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8713405132293701, "num_tokens": 296090591.0, "step": 7762 }, { "epoch": 0.98753339269813, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.347488403320312, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8584486246109009, "num_tokens": 296133024.0, "step": 7763 }, { "epoch": 0.9876606029767205, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.383367538452148, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8527935743331909, "num_tokens": 296177347.0, "step": 7764 }, { "epoch": 0.9877878132553111, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.341676712036133, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8547447919845581, "num_tokens": 296215532.0, "step": 7765 }, { "epoch": 0.9879150235339016, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.407506942749023, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8531965613365173, "num_tokens": 296249494.0, "step": 7766 }, { "epoch": 0.988042233812492, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.311562538146973, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8547917604446411, "num_tokens": 296289370.0, "step": 7767 }, { "epoch": 0.9881694440910825, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.382162094116211, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8653478622436523, "num_tokens": 296329376.0, "step": 7768 }, { "epoch": 0.9882966543696731, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.360300064086914, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8709505200386047, "num_tokens": 296368875.0, "step": 7769 }, { "epoch": 0.9884238646482636, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.386734008789062, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8625860214233398, "num_tokens": 296404766.0, "step": 7770 }, { "epoch": 0.9885510749268541, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.352230072021484, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8633517026901245, "num_tokens": 296441705.0, "step": 7771 }, { "epoch": 0.9886782852054447, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.357804298400879, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8705751895904541, "num_tokens": 296481075.0, "step": 7772 }, { "epoch": 0.9888054954840351, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.350607872009277, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8597555756568909, "num_tokens": 296511404.0, "step": 7773 }, { "epoch": 0.9889327057626256, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.359352111816406, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8549721240997314, "num_tokens": 296546080.0, "step": 7774 }, { "epoch": 0.9890599160412161, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.373859405517578, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8660777807235718, "num_tokens": 296586119.0, "step": 7775 }, { "epoch": 0.9891871263198067, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.319706916809082, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8629514575004578, "num_tokens": 296629602.0, "step": 7776 }, { "epoch": 0.9893143365983972, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.334739685058594, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8576652407646179, "num_tokens": 296666709.0, "step": 7777 }, { "epoch": 0.9894415468769877, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.326075553894043, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8639668822288513, "num_tokens": 296702611.0, "step": 7778 }, { "epoch": 0.9895687571555781, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.336000442504883, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8689793944358826, "num_tokens": 296737649.0, "step": 7779 }, { "epoch": 0.9896959674341687, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.32148265838623, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8660876154899597, "num_tokens": 296776915.0, "step": 7780 }, { "epoch": 0.9898231777127592, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.30618667602539, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8701179027557373, "num_tokens": 296811491.0, "step": 7781 }, { "epoch": 0.9899503879913497, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.364934921264648, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8677053451538086, "num_tokens": 296847602.0, "step": 7782 }, { "epoch": 0.9900775982699402, "ewc_loss": 0.024169921875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.334547996520996, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8475949168205261, "num_tokens": 296884737.0, "step": 7783 }, { "epoch": 0.9902048085485308, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.330909729003906, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8510345220565796, "num_tokens": 296927780.0, "step": 7784 }, { "epoch": 0.9903320188271212, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.332585334777832, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8779883980751038, "num_tokens": 296964200.0, "step": 7785 }, { "epoch": 0.9904592291057117, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.343486785888672, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8554037809371948, "num_tokens": 297009927.0, "step": 7786 }, { "epoch": 0.9905864393843022, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.359780311584473, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8516227006912231, "num_tokens": 297050788.0, "step": 7787 }, { "epoch": 0.9907136496628928, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.308931350708008, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8699436187744141, "num_tokens": 297092398.0, "step": 7788 }, { "epoch": 0.9908408599414833, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.373452186584473, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8623816967010498, "num_tokens": 297130728.0, "step": 7789 }, { "epoch": 0.9909680702200738, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.328509330749512, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8866955637931824, "num_tokens": 297169124.0, "step": 7790 }, { "epoch": 0.9910952804986642, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.357145309448242, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8734638094902039, "num_tokens": 297204914.0, "step": 7791 }, { "epoch": 0.9912224907772548, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.37165355682373, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8607631921768188, "num_tokens": 297240551.0, "step": 7792 }, { "epoch": 0.9913497010558453, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.372633934020996, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8553341031074524, "num_tokens": 297287160.0, "step": 7793 }, { "epoch": 0.9914769113344358, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.369487762451172, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8657073974609375, "num_tokens": 297323394.0, "step": 7794 }, { "epoch": 0.9916041216130264, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.402188301086426, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.8417485952377319, "num_tokens": 297357808.0, "step": 7795 }, { "epoch": 0.9917313318916169, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.362059593200684, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8585761189460754, "num_tokens": 297397300.0, "step": 7796 }, { "epoch": 0.9918585421702073, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.337964057922363, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8732078671455383, "num_tokens": 297434699.0, "step": 7797 }, { "epoch": 0.9919857524487978, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.382134437561035, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8631728291511536, "num_tokens": 297473939.0, "step": 7798 }, { "epoch": 0.9921129627273884, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.344635963439941, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8556391000747681, "num_tokens": 297511154.0, "step": 7799 }, { "epoch": 0.9922401730059789, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.356622695922852, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8551453351974487, "num_tokens": 297556092.0, "step": 7800 }, { "epoch": 0.9923673832845694, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.340127944946289, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8681902289390564, "num_tokens": 297598160.0, "step": 7801 }, { "epoch": 0.9924945935631599, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.32485580444336, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8800848722457886, "num_tokens": 297638706.0, "step": 7802 }, { "epoch": 0.9926218038417504, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.354737281799316, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8681108951568604, "num_tokens": 297675955.0, "step": 7803 }, { "epoch": 0.9927490141203409, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.323161125183105, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8593019247055054, "num_tokens": 297715362.0, "step": 7804 }, { "epoch": 0.9928762243989314, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.341283798217773, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.849027156829834, "num_tokens": 297750931.0, "step": 7805 }, { "epoch": 0.993003434677522, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.333579063415527, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8558406829833984, "num_tokens": 297786107.0, "step": 7806 }, { "epoch": 0.9931306449561125, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.347270965576172, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8635230660438538, "num_tokens": 297824782.0, "step": 7807 }, { "epoch": 0.993257855234703, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.311573028564453, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.859178364276886, "num_tokens": 297861748.0, "step": 7808 }, { "epoch": 0.9933850655132934, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.34067440032959, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.863890528678894, "num_tokens": 297902722.0, "step": 7809 }, { "epoch": 0.993512275791884, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.30614948272705, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8649675846099854, "num_tokens": 297948709.0, "step": 7810 }, { "epoch": 0.9936394860704745, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.36294937133789, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8651200532913208, "num_tokens": 297986446.0, "step": 7811 }, { "epoch": 0.993766696349065, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.326638221740723, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8555523753166199, "num_tokens": 298024004.0, "step": 7812 }, { "epoch": 0.9938939066276555, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.364306449890137, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8541358709335327, "num_tokens": 298065518.0, "step": 7813 }, { "epoch": 0.9940211169062461, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.332852363586426, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8561499118804932, "num_tokens": 298108138.0, "step": 7814 }, { "epoch": 0.9941483271848366, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.311185836791992, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8636016845703125, "num_tokens": 298149348.0, "step": 7815 }, { "epoch": 0.994275537463427, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.329086303710938, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8611412048339844, "num_tokens": 298186347.0, "step": 7816 }, { "epoch": 0.9944027477420175, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.38264274597168, "learning_rate": 1e-06, "loss": 0.4831, "mean_token_accuracy": 0.8421781659126282, "num_tokens": 298223322.0, "step": 7817 }, { "epoch": 0.9945299580206081, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.296306610107422, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.859320878982544, "num_tokens": 298258794.0, "step": 7818 }, { "epoch": 0.9946571682991986, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.35315227508545, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.857388436794281, "num_tokens": 298297510.0, "step": 7819 }, { "epoch": 0.9947843785777891, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.33178424835205, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8685787916183472, "num_tokens": 298333784.0, "step": 7820 }, { "epoch": 0.9949115888563796, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.326003074645996, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8700301647186279, "num_tokens": 298371241.0, "step": 7821 }, { "epoch": 0.9950387991349701, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.304436683654785, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8499147295951843, "num_tokens": 298407902.0, "step": 7822 }, { "epoch": 0.9951660094135606, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.320162773132324, "learning_rate": 1e-06, "loss": 0.502, "mean_token_accuracy": 0.8394633531570435, "num_tokens": 298452339.0, "step": 7823 }, { "epoch": 0.9952932196921511, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.297484397888184, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8538129329681396, "num_tokens": 298495854.0, "step": 7824 }, { "epoch": 0.9954204299707416, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.344613075256348, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8568571209907532, "num_tokens": 298529642.0, "step": 7825 }, { "epoch": 0.9955476402493322, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.324585914611816, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8645995855331421, "num_tokens": 298574396.0, "step": 7826 }, { "epoch": 0.9956748505279227, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.360525131225586, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8495837450027466, "num_tokens": 298612457.0, "step": 7827 }, { "epoch": 0.9958020608065131, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.361167907714844, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8632446527481079, "num_tokens": 298645061.0, "step": 7828 }, { "epoch": 0.9959292710851037, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.326165199279785, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8609785437583923, "num_tokens": 298688014.0, "step": 7829 }, { "epoch": 0.9960564813636942, "ewc_loss": 0.02392578125, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.312934875488281, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8659439086914062, "num_tokens": 298727815.0, "step": 7830 }, { "epoch": 0.9961836916422847, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.333259582519531, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8838127851486206, "num_tokens": 298765663.0, "step": 7831 }, { "epoch": 0.9963109019208752, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.35758113861084, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8665192127227783, "num_tokens": 298799360.0, "step": 7832 }, { "epoch": 0.9964381121994658, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.343645095825195, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8623632192611694, "num_tokens": 298835669.0, "step": 7833 }, { "epoch": 0.9965653224780562, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.373815536499023, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8608509302139282, "num_tokens": 298866871.0, "step": 7834 }, { "epoch": 0.9966925327566467, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.368949890136719, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8543928861618042, "num_tokens": 298904136.0, "step": 7835 }, { "epoch": 0.9968197430352372, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.348237037658691, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8757133483886719, "num_tokens": 298937069.0, "step": 7836 }, { "epoch": 0.9969469533138278, "ewc_loss": 0.024169921875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.377161979675293, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8552484512329102, "num_tokens": 298976778.0, "step": 7837 }, { "epoch": 0.9970741635924183, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.366737365722656, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8726059198379517, "num_tokens": 299010808.0, "step": 7838 }, { "epoch": 0.9972013738710088, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.393669128417969, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8517245650291443, "num_tokens": 299047611.0, "step": 7839 }, { "epoch": 0.9973285841495992, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.381816864013672, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8677670955657959, "num_tokens": 299088668.0, "step": 7840 }, { "epoch": 0.9974557944281898, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.362639427185059, "learning_rate": 1e-06, "loss": 0.4878, "mean_token_accuracy": 0.8430099487304688, "num_tokens": 299126572.0, "step": 7841 }, { "epoch": 0.9975830047067803, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.400002479553223, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8750730752944946, "num_tokens": 299164806.0, "step": 7842 }, { "epoch": 0.9977102149853708, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.38131332397461, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8586264252662659, "num_tokens": 299207039.0, "step": 7843 }, { "epoch": 0.9978374252639614, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.421985626220703, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8713867664337158, "num_tokens": 299243242.0, "step": 7844 }, { "epoch": 0.9979646355425519, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.351487159729004, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8695231676101685, "num_tokens": 299282326.0, "step": 7845 }, { "epoch": 0.9980918458211423, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.414515495300293, "learning_rate": 1e-06, "loss": 0.3324, "mean_token_accuracy": 0.8913226127624512, "num_tokens": 299316645.0, "step": 7846 }, { "epoch": 0.9982190560997328, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.434403419494629, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8653949499130249, "num_tokens": 299349932.0, "step": 7847 }, { "epoch": 0.9983462663783234, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.390908241271973, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8421056270599365, "num_tokens": 299392452.0, "step": 7848 }, { "epoch": 0.9984734766569139, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.398713111877441, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8687379360198975, "num_tokens": 299431479.0, "step": 7849 }, { "epoch": 0.9986006869355044, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.317651748657227, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8610019087791443, "num_tokens": 299475585.0, "step": 7850 }, { "epoch": 0.9987278972140949, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.4285249710083, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8788286447525024, "num_tokens": 299514026.0, "step": 7851 }, { "epoch": 0.9988551074926854, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.35562801361084, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8526533842086792, "num_tokens": 299550858.0, "step": 7852 }, { "epoch": 0.9989823177712759, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.406111717224121, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8559378385543823, "num_tokens": 299586831.0, "step": 7853 }, { "epoch": 0.9991095280498664, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.344738960266113, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.873257577419281, "num_tokens": 299623116.0, "step": 7854 }, { "epoch": 0.9992367383284569, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.38149642944336, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8668662309646606, "num_tokens": 299662165.0, "step": 7855 }, { "epoch": 0.9993639486070475, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.404892921447754, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8606283664703369, "num_tokens": 299702581.0, "step": 7856 }, { "epoch": 0.999491158885638, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.356588363647461, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8487545251846313, "num_tokens": 299742559.0, "step": 7857 }, { "epoch": 0.9996183691642284, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.409181594848633, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8794801831245422, "num_tokens": 299778637.0, "step": 7858 }, { "epoch": 0.9997455794428189, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.401023864746094, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8719346523284912, "num_tokens": 299812808.0, "step": 7859 }, { "epoch": 0.9998727897214095, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.392547607421875, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8596839904785156, "num_tokens": 299848987.0, "step": 7860 }, { "epoch": 1.0, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.32135009765625, "learning_rate": 1e-06, "loss": 0.4977, "mean_token_accuracy": 0.8447904586791992, "num_tokens": 299886286.0, "step": 7861 }, { "epoch": 1.0001272102785905, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.424243927001953, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8579366207122803, "num_tokens": 299925456.0, "step": 7862 }, { "epoch": 1.000254420557181, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.39218521118164, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8615397810935974, "num_tokens": 299965936.0, "step": 7863 }, { "epoch": 1.0003816308357716, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.346396446228027, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8772183656692505, "num_tokens": 300003181.0, "step": 7864 }, { "epoch": 1.0005088411143621, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.379629135131836, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8623396754264832, "num_tokens": 300040502.0, "step": 7865 }, { "epoch": 1.0006360513929526, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.374017715454102, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.879889965057373, "num_tokens": 300078795.0, "step": 7866 }, { "epoch": 1.0007632616715432, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.399879455566406, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8750365376472473, "num_tokens": 300115188.0, "step": 7867 }, { "epoch": 1.0008904719501335, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.38351058959961, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8669975996017456, "num_tokens": 300158583.0, "step": 7868 }, { "epoch": 1.001017682228724, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.425537109375, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8697899580001831, "num_tokens": 300194415.0, "step": 7869 }, { "epoch": 1.0011448925073145, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.397529602050781, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.860226035118103, "num_tokens": 300235066.0, "step": 7870 }, { "epoch": 1.001272102785905, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.433959007263184, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8675503730773926, "num_tokens": 300272167.0, "step": 7871 }, { "epoch": 1.0013993130644956, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 11.377907752990723, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8570324778556824, "num_tokens": 300306032.0, "step": 7872 }, { "epoch": 1.0015265233430861, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.417400360107422, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8624907732009888, "num_tokens": 300338433.0, "step": 7873 }, { "epoch": 1.0016537336216766, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.416451454162598, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8556502461433411, "num_tokens": 300376527.0, "step": 7874 }, { "epoch": 1.0017809439002672, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.353764533996582, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8824036121368408, "num_tokens": 300417309.0, "step": 7875 }, { "epoch": 1.0019081541788577, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.411564826965332, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8673335313796997, "num_tokens": 300459438.0, "step": 7876 }, { "epoch": 1.0020353644574482, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.397943496704102, "learning_rate": 1e-06, "loss": 0.4909, "mean_token_accuracy": 0.8424378633499146, "num_tokens": 300501443.0, "step": 7877 }, { "epoch": 1.0021625747360388, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.39535140991211, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.873284101486206, "num_tokens": 300531918.0, "step": 7878 }, { "epoch": 1.0022897850146293, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.380704879760742, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8596425652503967, "num_tokens": 300575084.0, "step": 7879 }, { "epoch": 1.0024169952932196, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.430411338806152, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8660764694213867, "num_tokens": 300608737.0, "step": 7880 }, { "epoch": 1.0025442055718101, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.396686553955078, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8536655902862549, "num_tokens": 300647855.0, "step": 7881 }, { "epoch": 1.0026714158504006, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.39228630065918, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8658458590507507, "num_tokens": 300689763.0, "step": 7882 }, { "epoch": 1.0027986261289912, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.374996185302734, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8756412267684937, "num_tokens": 300725838.0, "step": 7883 }, { "epoch": 1.0029258364075817, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.412639617919922, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8703885078430176, "num_tokens": 300760423.0, "step": 7884 }, { "epoch": 1.0030530466861722, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.395881652832031, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8719366788864136, "num_tokens": 300796236.0, "step": 7885 }, { "epoch": 1.0031802569647628, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.345423698425293, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8698626160621643, "num_tokens": 300835213.0, "step": 7886 }, { "epoch": 1.0033074672433533, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.412613868713379, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8718214631080627, "num_tokens": 300873363.0, "step": 7887 }, { "epoch": 1.0034346775219438, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.424786567687988, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8603402972221375, "num_tokens": 300910233.0, "step": 7888 }, { "epoch": 1.0035618878005343, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.416423797607422, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8726581931114197, "num_tokens": 300946986.0, "step": 7889 }, { "epoch": 1.0036890980791249, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.39920425415039, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8680088520050049, "num_tokens": 300985750.0, "step": 7890 }, { "epoch": 1.0038163083577154, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.403794288635254, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8645075559616089, "num_tokens": 301032820.0, "step": 7891 }, { "epoch": 1.0039435186363057, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.38860034942627, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8559001684188843, "num_tokens": 301070825.0, "step": 7892 }, { "epoch": 1.0040707289148962, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.360082626342773, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.861939013004303, "num_tokens": 301110764.0, "step": 7893 }, { "epoch": 1.0041979391934868, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.376654624938965, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8684201836585999, "num_tokens": 301152421.0, "step": 7894 }, { "epoch": 1.0043251494720773, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.367578506469727, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8665972948074341, "num_tokens": 301192780.0, "step": 7895 }, { "epoch": 1.0044523597506678, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.432890892028809, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8620772361755371, "num_tokens": 301228147.0, "step": 7896 }, { "epoch": 1.0045795700292584, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.407232284545898, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8755820989608765, "num_tokens": 301259568.0, "step": 7897 }, { "epoch": 1.0047067803078489, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.38293170928955, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8664849996566772, "num_tokens": 301296483.0, "step": 7898 }, { "epoch": 1.0048339905864394, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.453580856323242, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8688715696334839, "num_tokens": 301337437.0, "step": 7899 }, { "epoch": 1.00496120086503, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.389378547668457, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8650838136672974, "num_tokens": 301368777.0, "step": 7900 }, { "epoch": 1.0050884111436205, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.42795181274414, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8622894883155823, "num_tokens": 301404885.0, "step": 7901 }, { "epoch": 1.005215621422211, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.42577075958252, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8553706407546997, "num_tokens": 301445308.0, "step": 7902 }, { "epoch": 1.0053428317008015, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.425308227539062, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8620918989181519, "num_tokens": 301484845.0, "step": 7903 }, { "epoch": 1.0054700419793918, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.391005516052246, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8629573583602905, "num_tokens": 301524498.0, "step": 7904 }, { "epoch": 1.0055972522579824, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.392732620239258, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.870832085609436, "num_tokens": 301565522.0, "step": 7905 }, { "epoch": 1.0057244625365729, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.395620346069336, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8609689474105835, "num_tokens": 301603724.0, "step": 7906 }, { "epoch": 1.0058516728151634, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.437338829040527, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8622697591781616, "num_tokens": 301636650.0, "step": 7907 }, { "epoch": 1.005978883093754, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.40063190460205, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8556643128395081, "num_tokens": 301675328.0, "step": 7908 }, { "epoch": 1.0061060933723445, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.371641159057617, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8589786291122437, "num_tokens": 301710058.0, "step": 7909 }, { "epoch": 1.006233303650935, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.355971336364746, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8593308329582214, "num_tokens": 301753567.0, "step": 7910 }, { "epoch": 1.0063605139295255, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.348040580749512, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8726048469543457, "num_tokens": 301792895.0, "step": 7911 }, { "epoch": 1.006487724208116, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.39293384552002, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8759945631027222, "num_tokens": 301833001.0, "step": 7912 }, { "epoch": 1.0066149344867066, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.347638130187988, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8689420223236084, "num_tokens": 301871623.0, "step": 7913 }, { "epoch": 1.006742144765297, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.409770011901855, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8769282102584839, "num_tokens": 301903750.0, "step": 7914 }, { "epoch": 1.0068693550438876, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.317341804504395, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8667316436767578, "num_tokens": 301943079.0, "step": 7915 }, { "epoch": 1.0069965653224782, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.407088279724121, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8697149157524109, "num_tokens": 301980031.0, "step": 7916 }, { "epoch": 1.0071237756010685, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.355583190917969, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.865338921546936, "num_tokens": 302020454.0, "step": 7917 }, { "epoch": 1.007250985879659, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.368459701538086, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8601593971252441, "num_tokens": 302064391.0, "step": 7918 }, { "epoch": 1.0073781961582495, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.378284454345703, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8653591275215149, "num_tokens": 302101598.0, "step": 7919 }, { "epoch": 1.00750540643684, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.322871208190918, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8690794706344604, "num_tokens": 302138803.0, "step": 7920 }, { "epoch": 1.0076326167154306, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.468498229980469, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8634476661682129, "num_tokens": 302173682.0, "step": 7921 }, { "epoch": 1.0077598269940211, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.353170394897461, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8612484931945801, "num_tokens": 302207390.0, "step": 7922 }, { "epoch": 1.0078870372726116, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.382946014404297, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8714213371276855, "num_tokens": 302244440.0, "step": 7923 }, { "epoch": 1.0080142475512022, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.386515617370605, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8604534864425659, "num_tokens": 302281058.0, "step": 7924 }, { "epoch": 1.0081414578297927, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.380227088928223, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.881717324256897, "num_tokens": 302319586.0, "step": 7925 }, { "epoch": 1.0082686681083832, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.379352569580078, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8594231009483337, "num_tokens": 302361356.0, "step": 7926 }, { "epoch": 1.0083958783869738, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.416487693786621, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8745357990264893, "num_tokens": 302399974.0, "step": 7927 }, { "epoch": 1.0085230886655643, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.408843994140625, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8678532838821411, "num_tokens": 302435709.0, "step": 7928 }, { "epoch": 1.0086502989441546, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.456815719604492, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8613426089286804, "num_tokens": 302474151.0, "step": 7929 }, { "epoch": 1.0087775092227451, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.45489501953125, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8677684664726257, "num_tokens": 302509877.0, "step": 7930 }, { "epoch": 1.0089047195013356, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.415332794189453, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8788366317749023, "num_tokens": 302547689.0, "step": 7931 }, { "epoch": 1.0090319297799262, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.385588645935059, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8635247945785522, "num_tokens": 302590728.0, "step": 7932 }, { "epoch": 1.0091591400585167, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.434418678283691, "learning_rate": 1e-06, "loss": 0.4921, "mean_token_accuracy": 0.8431256413459778, "num_tokens": 302625709.0, "step": 7933 }, { "epoch": 1.0092863503371072, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.459303855895996, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8738238215446472, "num_tokens": 302662086.0, "step": 7934 }, { "epoch": 1.0094135606156978, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.399934768676758, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8819448351860046, "num_tokens": 302703680.0, "step": 7935 }, { "epoch": 1.0095407708942883, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.411543846130371, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8511439561843872, "num_tokens": 302740426.0, "step": 7936 }, { "epoch": 1.0096679811728788, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.41047477722168, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.878407895565033, "num_tokens": 302780778.0, "step": 7937 }, { "epoch": 1.0097951914514693, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.40073013305664, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.868977427482605, "num_tokens": 302816012.0, "step": 7938 }, { "epoch": 1.0099224017300599, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.422309875488281, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8442087173461914, "num_tokens": 302858325.0, "step": 7939 }, { "epoch": 1.0100496120086504, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.397045135498047, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8669587969779968, "num_tokens": 302898210.0, "step": 7940 }, { "epoch": 1.0101768222872407, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.353598594665527, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8657388687133789, "num_tokens": 302936072.0, "step": 7941 }, { "epoch": 1.0103040325658312, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.421514511108398, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8566920757293701, "num_tokens": 302978941.0, "step": 7942 }, { "epoch": 1.0104312428444218, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.393787384033203, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8577982187271118, "num_tokens": 303016664.0, "step": 7943 }, { "epoch": 1.0105584531230123, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.375829696655273, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8616317510604858, "num_tokens": 303063313.0, "step": 7944 }, { "epoch": 1.0106856634016028, "ewc_loss": 0.02392578125, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.360250473022461, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8554019331932068, "num_tokens": 303103786.0, "step": 7945 }, { "epoch": 1.0108128736801933, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.431281089782715, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8706105947494507, "num_tokens": 303136928.0, "step": 7946 }, { "epoch": 1.0109400839587839, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.364699363708496, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8728123903274536, "num_tokens": 303174854.0, "step": 7947 }, { "epoch": 1.0110672942373744, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.392805099487305, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8634136915206909, "num_tokens": 303214234.0, "step": 7948 }, { "epoch": 1.011194504515965, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.385720252990723, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8787966370582581, "num_tokens": 303251830.0, "step": 7949 }, { "epoch": 1.0113217147945555, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.401710510253906, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8553909659385681, "num_tokens": 303291564.0, "step": 7950 }, { "epoch": 1.011448925073146, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.434223175048828, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8628398776054382, "num_tokens": 303328642.0, "step": 7951 }, { "epoch": 1.0115761353517365, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.343381881713867, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8626246452331543, "num_tokens": 303373693.0, "step": 7952 }, { "epoch": 1.0117033456303268, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.404053688049316, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8690733909606934, "num_tokens": 303407124.0, "step": 7953 }, { "epoch": 1.0118305559089174, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.422327041625977, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8767620921134949, "num_tokens": 303441586.0, "step": 7954 }, { "epoch": 1.0119577661875079, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.411545753479004, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8665959239006042, "num_tokens": 303473593.0, "step": 7955 }, { "epoch": 1.0120849764660984, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.419417381286621, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8621668815612793, "num_tokens": 303510239.0, "step": 7956 }, { "epoch": 1.012212186744689, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.381071090698242, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8623444437980652, "num_tokens": 303544709.0, "step": 7957 }, { "epoch": 1.0123393970232795, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.379279136657715, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8744847178459167, "num_tokens": 303581990.0, "step": 7958 }, { "epoch": 1.01246660730187, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.38558292388916, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.868604838848114, "num_tokens": 303625167.0, "step": 7959 }, { "epoch": 1.0125938175804605, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.395163536071777, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8658583164215088, "num_tokens": 303664751.0, "step": 7960 }, { "epoch": 1.012721027859051, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.403164863586426, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8634858131408691, "num_tokens": 303702190.0, "step": 7961 }, { "epoch": 1.0128482381376416, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.318174362182617, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8796972632408142, "num_tokens": 303742872.0, "step": 7962 }, { "epoch": 1.012975448416232, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.399152755737305, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8574506044387817, "num_tokens": 303781265.0, "step": 7963 }, { "epoch": 1.0131026586948226, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.418275833129883, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8661047220230103, "num_tokens": 303820686.0, "step": 7964 }, { "epoch": 1.0132298689734132, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.391152381896973, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8685475587844849, "num_tokens": 303863428.0, "step": 7965 }, { "epoch": 1.0133570792520035, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.36677360534668, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8601745367050171, "num_tokens": 303903847.0, "step": 7966 }, { "epoch": 1.013484289530594, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.399147033691406, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8573639988899231, "num_tokens": 303943083.0, "step": 7967 }, { "epoch": 1.0136114998091845, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.400236129760742, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.875274121761322, "num_tokens": 303978773.0, "step": 7968 }, { "epoch": 1.013738710087775, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.450976371765137, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8729551434516907, "num_tokens": 304017058.0, "step": 7969 }, { "epoch": 1.0138659203663656, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.36005687713623, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8572155237197876, "num_tokens": 304054104.0, "step": 7970 }, { "epoch": 1.013993130644956, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.425073623657227, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8792203664779663, "num_tokens": 304091100.0, "step": 7971 }, { "epoch": 1.0141203409235466, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.350700378417969, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8660693764686584, "num_tokens": 304127747.0, "step": 7972 }, { "epoch": 1.0142475512021372, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 11.35041618347168, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8663699626922607, "num_tokens": 304171089.0, "step": 7973 }, { "epoch": 1.0143747614807277, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.364846229553223, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8563733100891113, "num_tokens": 304210207.0, "step": 7974 }, { "epoch": 1.0145019717593182, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.388100624084473, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8828567266464233, "num_tokens": 304249423.0, "step": 7975 }, { "epoch": 1.0146291820379088, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.340912818908691, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8695688247680664, "num_tokens": 304286872.0, "step": 7976 }, { "epoch": 1.0147563923164993, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.421607971191406, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.870148777961731, "num_tokens": 304323033.0, "step": 7977 }, { "epoch": 1.0148836025950896, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.437576293945312, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8694124221801758, "num_tokens": 304360096.0, "step": 7978 }, { "epoch": 1.0150108128736801, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.357768058776855, "learning_rate": 1e-06, "loss": 0.3451, "mean_token_accuracy": 0.8860206604003906, "num_tokens": 304396364.0, "step": 7979 }, { "epoch": 1.0151380231522706, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.369222640991211, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8676666617393494, "num_tokens": 304431583.0, "step": 7980 }, { "epoch": 1.0152652334308612, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.403535842895508, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8609337210655212, "num_tokens": 304470531.0, "step": 7981 }, { "epoch": 1.0153924437094517, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.398093223571777, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8707855939865112, "num_tokens": 304505187.0, "step": 7982 }, { "epoch": 1.0155196539880422, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.405881881713867, "learning_rate": 1e-06, "loss": 0.4852, "mean_token_accuracy": 0.8444404602050781, "num_tokens": 304542685.0, "step": 7983 }, { "epoch": 1.0156468642666328, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.390024185180664, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8663755059242249, "num_tokens": 304579204.0, "step": 7984 }, { "epoch": 1.0157740745452233, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.37491512298584, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8734837174415588, "num_tokens": 304615224.0, "step": 7985 }, { "epoch": 1.0159012848238138, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.439982414245605, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8614650964736938, "num_tokens": 304653513.0, "step": 7986 }, { "epoch": 1.0160284951024043, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.38473892211914, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8713135719299316, "num_tokens": 304686744.0, "step": 7987 }, { "epoch": 1.0161557053809949, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.387904167175293, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8655893206596375, "num_tokens": 304727262.0, "step": 7988 }, { "epoch": 1.0162829156595854, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.446372985839844, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8640580773353577, "num_tokens": 304763418.0, "step": 7989 }, { "epoch": 1.0164101259381757, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.357845306396484, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8694480061531067, "num_tokens": 304800763.0, "step": 7990 }, { "epoch": 1.0165373362167662, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.43472957611084, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.884537935256958, "num_tokens": 304833136.0, "step": 7991 }, { "epoch": 1.0166645464953568, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.393546104431152, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8759698867797852, "num_tokens": 304873208.0, "step": 7992 }, { "epoch": 1.0167917567739473, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.460639953613281, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8674007654190063, "num_tokens": 304914237.0, "step": 7993 }, { "epoch": 1.0169189670525378, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.357186317443848, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8816850185394287, "num_tokens": 304955453.0, "step": 7994 }, { "epoch": 1.0170461773311283, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.419493675231934, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8661051988601685, "num_tokens": 304990533.0, "step": 7995 }, { "epoch": 1.0171733876097189, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.453936576843262, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8655447959899902, "num_tokens": 305024849.0, "step": 7996 }, { "epoch": 1.0173005978883094, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.418256759643555, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8633304238319397, "num_tokens": 305061615.0, "step": 7997 }, { "epoch": 1.0174278081669, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.42343521118164, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8473155498504639, "num_tokens": 305102251.0, "step": 7998 }, { "epoch": 1.0175550184454905, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.409982681274414, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8754843473434448, "num_tokens": 305135405.0, "step": 7999 }, { "epoch": 1.017682228724081, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.430862426757812, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.865270733833313, "num_tokens": 305169005.0, "step": 8000 }, { "epoch": 1.0178094390026715, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.408281326293945, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8626459240913391, "num_tokens": 305209050.0, "step": 8001 }, { "epoch": 1.0179366492812618, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.444557189941406, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8667705059051514, "num_tokens": 305238250.0, "step": 8002 }, { "epoch": 1.0180638595598523, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.439934730529785, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8696531653404236, "num_tokens": 305283544.0, "step": 8003 }, { "epoch": 1.0181910698384429, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.432181358337402, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8499080538749695, "num_tokens": 305320548.0, "step": 8004 }, { "epoch": 1.0183182801170334, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.382628440856934, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.87418532371521, "num_tokens": 305359560.0, "step": 8005 }, { "epoch": 1.018445490395624, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.4674072265625, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8632968664169312, "num_tokens": 305396680.0, "step": 8006 }, { "epoch": 1.0185727006742145, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.410161972045898, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8753877878189087, "num_tokens": 305432856.0, "step": 8007 }, { "epoch": 1.018699910952805, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.438364028930664, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8683440685272217, "num_tokens": 305469973.0, "step": 8008 }, { "epoch": 1.0188271212313955, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.450285911560059, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8726813793182373, "num_tokens": 305505576.0, "step": 8009 }, { "epoch": 1.018954331509986, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.435263633728027, "learning_rate": 1e-06, "loss": 0.4953, "mean_token_accuracy": 0.8482177257537842, "num_tokens": 305542037.0, "step": 8010 }, { "epoch": 1.0190815417885766, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.40422534942627, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8737787008285522, "num_tokens": 305587784.0, "step": 8011 }, { "epoch": 1.019208752067167, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.470842361450195, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8586726188659668, "num_tokens": 305621888.0, "step": 8012 }, { "epoch": 1.0193359623457576, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.399155616760254, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8863735795021057, "num_tokens": 305663844.0, "step": 8013 }, { "epoch": 1.0194631726243482, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.428239822387695, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.881314218044281, "num_tokens": 305695913.0, "step": 8014 }, { "epoch": 1.0195903829029385, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.434881210327148, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8650462031364441, "num_tokens": 305733632.0, "step": 8015 }, { "epoch": 1.019717593181529, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.497730255126953, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8512499332427979, "num_tokens": 305766917.0, "step": 8016 }, { "epoch": 1.0198448034601195, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.432317733764648, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8609969019889832, "num_tokens": 305805768.0, "step": 8017 }, { "epoch": 1.01997201373871, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.462726593017578, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8668935894966125, "num_tokens": 305843412.0, "step": 8018 }, { "epoch": 1.0200992240173006, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.431492805480957, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8818681836128235, "num_tokens": 305880523.0, "step": 8019 }, { "epoch": 1.020226434295891, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.46056842803955, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8689306974411011, "num_tokens": 305915379.0, "step": 8020 }, { "epoch": 1.0203536445744816, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.491329193115234, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8813390135765076, "num_tokens": 305952471.0, "step": 8021 }, { "epoch": 1.0204808548530722, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.429130554199219, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8658942580223083, "num_tokens": 305994023.0, "step": 8022 }, { "epoch": 1.0206080651316627, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.466575622558594, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8715354800224304, "num_tokens": 306026563.0, "step": 8023 }, { "epoch": 1.0207352754102532, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.407308578491211, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8636454939842224, "num_tokens": 306067913.0, "step": 8024 }, { "epoch": 1.0208624856888437, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.449211120605469, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.863983154296875, "num_tokens": 306108692.0, "step": 8025 }, { "epoch": 1.0209896959674343, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.421090126037598, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8678940534591675, "num_tokens": 306148082.0, "step": 8026 }, { "epoch": 1.0211169062460246, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.422335624694824, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8742029666900635, "num_tokens": 306182800.0, "step": 8027 }, { "epoch": 1.021244116524615, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.407770156860352, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.863756000995636, "num_tokens": 306226741.0, "step": 8028 }, { "epoch": 1.0213713268032056, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.409688949584961, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8651025891304016, "num_tokens": 306265993.0, "step": 8029 }, { "epoch": 1.0214985370817962, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.403535842895508, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8705788850784302, "num_tokens": 306306511.0, "step": 8030 }, { "epoch": 1.0216257473603867, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.42801570892334, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8810027837753296, "num_tokens": 306337687.0, "step": 8031 }, { "epoch": 1.0217529576389772, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.417311668395996, "learning_rate": 1e-06, "loss": 0.5137, "mean_token_accuracy": 0.8325303792953491, "num_tokens": 306379646.0, "step": 8032 }, { "epoch": 1.0218801679175677, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.433497428894043, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8669888973236084, "num_tokens": 306415389.0, "step": 8033 }, { "epoch": 1.0220073781961583, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.419286727905273, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8623486161231995, "num_tokens": 306452035.0, "step": 8034 }, { "epoch": 1.0221345884747488, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.419437408447266, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8660383820533752, "num_tokens": 306496277.0, "step": 8035 }, { "epoch": 1.0222617987533393, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.476451873779297, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8721165657043457, "num_tokens": 306528212.0, "step": 8036 }, { "epoch": 1.0223890090319299, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.457736015319824, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8712942600250244, "num_tokens": 306562047.0, "step": 8037 }, { "epoch": 1.0225162193105204, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.49824333190918, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8572812080383301, "num_tokens": 306601298.0, "step": 8038 }, { "epoch": 1.0226434295891107, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.430581092834473, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8661245107650757, "num_tokens": 306640961.0, "step": 8039 }, { "epoch": 1.0227706398677012, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.403996467590332, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8846905827522278, "num_tokens": 306679507.0, "step": 8040 }, { "epoch": 1.0228978501462918, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.422557830810547, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8577741384506226, "num_tokens": 306717337.0, "step": 8041 }, { "epoch": 1.0230250604248823, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.455520629882812, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.841629147529602, "num_tokens": 306754152.0, "step": 8042 }, { "epoch": 1.0231522707034728, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.4446382522583, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8596448302268982, "num_tokens": 306794887.0, "step": 8043 }, { "epoch": 1.0232794809820633, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.409780502319336, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8695544004440308, "num_tokens": 306828374.0, "step": 8044 }, { "epoch": 1.0234066912606539, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.42659854888916, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8708170652389526, "num_tokens": 306865285.0, "step": 8045 }, { "epoch": 1.0235339015392444, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.433623313903809, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.872836172580719, "num_tokens": 306897195.0, "step": 8046 }, { "epoch": 1.023661111817835, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.448978424072266, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.865312933921814, "num_tokens": 306935352.0, "step": 8047 }, { "epoch": 1.0237883220964255, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.480645179748535, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8721959590911865, "num_tokens": 306964621.0, "step": 8048 }, { "epoch": 1.023915532375016, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.474654197692871, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.865291178226471, "num_tokens": 307003451.0, "step": 8049 }, { "epoch": 1.0240427426536065, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.376044273376465, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8662549257278442, "num_tokens": 307044695.0, "step": 8050 }, { "epoch": 1.0241699529321968, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.474056243896484, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8738123178482056, "num_tokens": 307080603.0, "step": 8051 }, { "epoch": 1.0242971632107873, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.41198444366455, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8600849509239197, "num_tokens": 307122300.0, "step": 8052 }, { "epoch": 1.0244243734893779, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.437253952026367, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8725986480712891, "num_tokens": 307158685.0, "step": 8053 }, { "epoch": 1.0245515837679684, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.433441162109375, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8723500370979309, "num_tokens": 307199104.0, "step": 8054 }, { "epoch": 1.024678794046559, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.449559211730957, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.858948826789856, "num_tokens": 307235840.0, "step": 8055 }, { "epoch": 1.0248060043251495, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.444489479064941, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8556843996047974, "num_tokens": 307273089.0, "step": 8056 }, { "epoch": 1.02493321460374, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.452898979187012, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8672103881835938, "num_tokens": 307318825.0, "step": 8057 }, { "epoch": 1.0250604248823305, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.420672416687012, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8736234903335571, "num_tokens": 307357006.0, "step": 8058 }, { "epoch": 1.025187635160921, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.463266372680664, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8650861382484436, "num_tokens": 307390274.0, "step": 8059 }, { "epoch": 1.0253148454395116, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.50048542022705, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8747565746307373, "num_tokens": 307427157.0, "step": 8060 }, { "epoch": 1.025442055718102, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.410698890686035, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8638349771499634, "num_tokens": 307466555.0, "step": 8061 }, { "epoch": 1.0255692659966926, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.460582733154297, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8592351078987122, "num_tokens": 307503375.0, "step": 8062 }, { "epoch": 1.0256964762752832, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.443209648132324, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8766167163848877, "num_tokens": 307543170.0, "step": 8063 }, { "epoch": 1.0258236865538735, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.404218673706055, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8635207414627075, "num_tokens": 307582890.0, "step": 8064 }, { "epoch": 1.025950896832464, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.493197441101074, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8598147630691528, "num_tokens": 307617643.0, "step": 8065 }, { "epoch": 1.0260781071110545, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.515754699707031, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8634347319602966, "num_tokens": 307652258.0, "step": 8066 }, { "epoch": 1.026205317389645, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.415404319763184, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8624467253684998, "num_tokens": 307693245.0, "step": 8067 }, { "epoch": 1.0263325276682356, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.416625022888184, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8691015839576721, "num_tokens": 307728751.0, "step": 8068 }, { "epoch": 1.026459737946826, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.518878936767578, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8558310270309448, "num_tokens": 307767869.0, "step": 8069 }, { "epoch": 1.0265869482254166, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.379810333251953, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8753786683082581, "num_tokens": 307809438.0, "step": 8070 }, { "epoch": 1.0267141585040072, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.482382774353027, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8470765352249146, "num_tokens": 307848752.0, "step": 8071 }, { "epoch": 1.0268413687825977, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.458639144897461, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8611596822738647, "num_tokens": 307886080.0, "step": 8072 }, { "epoch": 1.0269685790611882, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.474315643310547, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8800780177116394, "num_tokens": 307922854.0, "step": 8073 }, { "epoch": 1.0270957893397787, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.447731971740723, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8635696768760681, "num_tokens": 307960032.0, "step": 8074 }, { "epoch": 1.0272229996183693, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.489381790161133, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8658696413040161, "num_tokens": 307996074.0, "step": 8075 }, { "epoch": 1.0273502098969596, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.417322158813477, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8786630034446716, "num_tokens": 308030833.0, "step": 8076 }, { "epoch": 1.02747742017555, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.425870895385742, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8764921426773071, "num_tokens": 308079620.0, "step": 8077 }, { "epoch": 1.0276046304541406, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.436084747314453, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8601205945014954, "num_tokens": 308118555.0, "step": 8078 }, { "epoch": 1.0277318407327312, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.480011940002441, "learning_rate": 1e-06, "loss": 0.4933, "mean_token_accuracy": 0.8500164151191711, "num_tokens": 308154410.0, "step": 8079 }, { "epoch": 1.0278590510113217, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.472776412963867, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8667426705360413, "num_tokens": 308191990.0, "step": 8080 }, { "epoch": 1.0279862612899122, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.378266334533691, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8467485904693604, "num_tokens": 308231047.0, "step": 8081 }, { "epoch": 1.0281134715685027, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.445825576782227, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8806288242340088, "num_tokens": 308269487.0, "step": 8082 }, { "epoch": 1.0282406818470933, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.418148040771484, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8659173846244812, "num_tokens": 308304974.0, "step": 8083 }, { "epoch": 1.0283678921256838, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.388859748840332, "learning_rate": 1e-06, "loss": 0.346, "mean_token_accuracy": 0.888115644454956, "num_tokens": 308342072.0, "step": 8084 }, { "epoch": 1.0284951024042743, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.420666694641113, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8624047040939331, "num_tokens": 308374293.0, "step": 8085 }, { "epoch": 1.0286223126828649, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.410504341125488, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8711186647415161, "num_tokens": 308415872.0, "step": 8086 }, { "epoch": 1.0287495229614554, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.460752487182617, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8597593903541565, "num_tokens": 308451198.0, "step": 8087 }, { "epoch": 1.0288767332400457, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.456558227539062, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8741632699966431, "num_tokens": 308490554.0, "step": 8088 }, { "epoch": 1.0290039435186362, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.383447647094727, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.860764741897583, "num_tokens": 308530221.0, "step": 8089 }, { "epoch": 1.0291311537972267, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.46121883392334, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8510688543319702, "num_tokens": 308563634.0, "step": 8090 }, { "epoch": 1.0292583640758173, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.416345596313477, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8666130304336548, "num_tokens": 308605033.0, "step": 8091 }, { "epoch": 1.0293855743544078, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.390212059020996, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8665778040885925, "num_tokens": 308642603.0, "step": 8092 }, { "epoch": 1.0295127846329983, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.397743225097656, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8645424842834473, "num_tokens": 308685189.0, "step": 8093 }, { "epoch": 1.0296399949115889, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.463860511779785, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8663688898086548, "num_tokens": 308726052.0, "step": 8094 }, { "epoch": 1.0297672051901794, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.443161964416504, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8549413681030273, "num_tokens": 308764828.0, "step": 8095 }, { "epoch": 1.02989441546877, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.462042808532715, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8766207695007324, "num_tokens": 308799712.0, "step": 8096 }, { "epoch": 1.0300216257473604, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.412731170654297, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8665804266929626, "num_tokens": 308839879.0, "step": 8097 }, { "epoch": 1.030148836025951, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.542341232299805, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8594744205474854, "num_tokens": 308880362.0, "step": 8098 }, { "epoch": 1.0302760463045415, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.411287307739258, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8764262795448303, "num_tokens": 308921395.0, "step": 8099 }, { "epoch": 1.0304032565831318, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.502486228942871, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8746482729911804, "num_tokens": 308959976.0, "step": 8100 }, { "epoch": 1.0305304668617223, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.41018295288086, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8794816732406616, "num_tokens": 309000702.0, "step": 8101 }, { "epoch": 1.0306576771403129, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.481740951538086, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8842607140541077, "num_tokens": 309036540.0, "step": 8102 }, { "epoch": 1.0307848874189034, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.46511173248291, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8662196397781372, "num_tokens": 309072408.0, "step": 8103 }, { "epoch": 1.030912097697494, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.514154434204102, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8532180786132812, "num_tokens": 309110366.0, "step": 8104 }, { "epoch": 1.0310393079760845, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.391477584838867, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8699440956115723, "num_tokens": 309155194.0, "step": 8105 }, { "epoch": 1.031166518254675, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.457351684570312, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8755823373794556, "num_tokens": 309193747.0, "step": 8106 }, { "epoch": 1.0312937285332655, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.415366172790527, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8546469211578369, "num_tokens": 309234161.0, "step": 8107 }, { "epoch": 1.031420938811856, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.431354522705078, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8558322191238403, "num_tokens": 309272574.0, "step": 8108 }, { "epoch": 1.0315481490904466, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.43968391418457, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8693748712539673, "num_tokens": 309306567.0, "step": 8109 }, { "epoch": 1.031675359369037, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.45838451385498, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8711934685707092, "num_tokens": 309347218.0, "step": 8110 }, { "epoch": 1.0318025696476276, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.4142484664917, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8808516263961792, "num_tokens": 309385701.0, "step": 8111 }, { "epoch": 1.0319297799262181, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.494392395019531, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8649631142616272, "num_tokens": 309424600.0, "step": 8112 }, { "epoch": 1.0320569902048085, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.512699127197266, "learning_rate": 1e-06, "loss": 0.4799, "mean_token_accuracy": 0.8457762002944946, "num_tokens": 309459929.0, "step": 8113 }, { "epoch": 1.032184200483399, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.443187713623047, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8581598997116089, "num_tokens": 309505085.0, "step": 8114 }, { "epoch": 1.0323114107619895, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.502821922302246, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8686956167221069, "num_tokens": 309544746.0, "step": 8115 }, { "epoch": 1.03243862104058, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.446471214294434, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8538839221000671, "num_tokens": 309584301.0, "step": 8116 }, { "epoch": 1.0325658313191706, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.457745552062988, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8703744411468506, "num_tokens": 309622570.0, "step": 8117 }, { "epoch": 1.032693041597761, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.438508033752441, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8508527874946594, "num_tokens": 309666417.0, "step": 8118 }, { "epoch": 1.0328202518763516, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.453705787658691, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8568209409713745, "num_tokens": 309705706.0, "step": 8119 }, { "epoch": 1.0329474621549422, "ewc_loss": 0.024169921875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 11.484641075134277, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8756859302520752, "num_tokens": 309744288.0, "step": 8120 }, { "epoch": 1.0330746724335327, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.42109489440918, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8535873293876648, "num_tokens": 309780406.0, "step": 8121 }, { "epoch": 1.0332018827121232, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.543603897094727, "learning_rate": 1e-06, "loss": 0.5088, "mean_token_accuracy": 0.8345628976821899, "num_tokens": 309811135.0, "step": 8122 }, { "epoch": 1.0333290929907137, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.472841262817383, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8712317943572998, "num_tokens": 309847037.0, "step": 8123 }, { "epoch": 1.0334563032693043, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.403986930847168, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.867129385471344, "num_tokens": 309884274.0, "step": 8124 }, { "epoch": 1.0335835135478946, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.494738578796387, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8789749145507812, "num_tokens": 309915572.0, "step": 8125 }, { "epoch": 1.033710723826485, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.456981658935547, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8716280460357666, "num_tokens": 309954502.0, "step": 8126 }, { "epoch": 1.0338379341050756, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.344476699829102, "learning_rate": 1e-06, "loss": 0.3461, "mean_token_accuracy": 0.8878755569458008, "num_tokens": 309993015.0, "step": 8127 }, { "epoch": 1.0339651443836662, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.533537864685059, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8573426604270935, "num_tokens": 310030840.0, "step": 8128 }, { "epoch": 1.0340923546622567, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.444673538208008, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8665844798088074, "num_tokens": 310067586.0, "step": 8129 }, { "epoch": 1.0342195649408472, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.475916862487793, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8756189942359924, "num_tokens": 310104496.0, "step": 8130 }, { "epoch": 1.0343467752194377, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.486302375793457, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8498075008392334, "num_tokens": 310141162.0, "step": 8131 }, { "epoch": 1.0344739854980283, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.478507995605469, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8649325370788574, "num_tokens": 310182965.0, "step": 8132 }, { "epoch": 1.0346011957766188, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.450425148010254, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8494019508361816, "num_tokens": 310219572.0, "step": 8133 }, { "epoch": 1.0347284060552093, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.523791313171387, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8556911945343018, "num_tokens": 310248711.0, "step": 8134 }, { "epoch": 1.0348556163337999, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.42639446258545, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8497112393379211, "num_tokens": 310287915.0, "step": 8135 }, { "epoch": 1.0349828266123904, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.484522819519043, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.864707350730896, "num_tokens": 310325619.0, "step": 8136 }, { "epoch": 1.0351100368909807, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.500853538513184, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8733776807785034, "num_tokens": 310362456.0, "step": 8137 }, { "epoch": 1.0352372471695712, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.460596084594727, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.861656904220581, "num_tokens": 310399584.0, "step": 8138 }, { "epoch": 1.0353644574481617, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.482946395874023, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8602564334869385, "num_tokens": 310437435.0, "step": 8139 }, { "epoch": 1.0354916677267523, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.441057205200195, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8711520433425903, "num_tokens": 310476320.0, "step": 8140 }, { "epoch": 1.0356188780053428, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.450276374816895, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8814433813095093, "num_tokens": 310512956.0, "step": 8141 }, { "epoch": 1.0357460882839333, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.476737022399902, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8645303845405579, "num_tokens": 310547605.0, "step": 8142 }, { "epoch": 1.0358732985625239, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.490434646606445, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8642619848251343, "num_tokens": 310586991.0, "step": 8143 }, { "epoch": 1.0360005088411144, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.465143203735352, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8655346632003784, "num_tokens": 310621271.0, "step": 8144 }, { "epoch": 1.036127719119705, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.426836967468262, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8684000968933105, "num_tokens": 310662154.0, "step": 8145 }, { "epoch": 1.0362549293982954, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.451854705810547, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.876250684261322, "num_tokens": 310701924.0, "step": 8146 }, { "epoch": 1.036382139676886, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.49011516571045, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8715598583221436, "num_tokens": 310739688.0, "step": 8147 }, { "epoch": 1.0365093499554765, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.50190544128418, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8600202798843384, "num_tokens": 310773317.0, "step": 8148 }, { "epoch": 1.0366365602340668, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.453998565673828, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8642796277999878, "num_tokens": 310810854.0, "step": 8149 }, { "epoch": 1.0367637705126573, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.451020240783691, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8591889142990112, "num_tokens": 310846080.0, "step": 8150 }, { "epoch": 1.0368909807912479, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.529616355895996, "learning_rate": 1e-06, "loss": 0.4882, "mean_token_accuracy": 0.842842698097229, "num_tokens": 310885032.0, "step": 8151 }, { "epoch": 1.0370181910698384, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.410715103149414, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8718504905700684, "num_tokens": 310924518.0, "step": 8152 }, { "epoch": 1.037145401348429, "ewc_loss": 0.0244140625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 11.506634712219238, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8678604364395142, "num_tokens": 310961734.0, "step": 8153 }, { "epoch": 1.0372726116270194, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.396072387695312, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8685657978057861, "num_tokens": 311004541.0, "step": 8154 }, { "epoch": 1.03739982190561, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.48321533203125, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8743917942047119, "num_tokens": 311043947.0, "step": 8155 }, { "epoch": 1.0375270321842005, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.42661190032959, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.862311840057373, "num_tokens": 311083154.0, "step": 8156 }, { "epoch": 1.037654242462791, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.471096992492676, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8772560358047485, "num_tokens": 311121429.0, "step": 8157 }, { "epoch": 1.0377814527413816, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.450867652893066, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8508465886116028, "num_tokens": 311166578.0, "step": 8158 }, { "epoch": 1.037908663019972, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.453765869140625, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8688485622406006, "num_tokens": 311209645.0, "step": 8159 }, { "epoch": 1.0380358732985626, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.45930290222168, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8703515529632568, "num_tokens": 311248428.0, "step": 8160 }, { "epoch": 1.0381630835771531, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.446662902832031, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8673772811889648, "num_tokens": 311282040.0, "step": 8161 }, { "epoch": 1.0382902938557435, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.481733322143555, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8728380799293518, "num_tokens": 311320233.0, "step": 8162 }, { "epoch": 1.038417504134334, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.41700267791748, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8519083857536316, "num_tokens": 311357586.0, "step": 8163 }, { "epoch": 1.0385447144129245, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.377939224243164, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8656267523765564, "num_tokens": 311395545.0, "step": 8164 }, { "epoch": 1.038671924691515, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.473814964294434, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.865603506565094, "num_tokens": 311428785.0, "step": 8165 }, { "epoch": 1.0387991349701056, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.42448616027832, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8549084663391113, "num_tokens": 311469546.0, "step": 8166 }, { "epoch": 1.038926345248696, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.414667129516602, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8578787446022034, "num_tokens": 311504064.0, "step": 8167 }, { "epoch": 1.0390535555272866, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.490344047546387, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8486838340759277, "num_tokens": 311536600.0, "step": 8168 }, { "epoch": 1.0391807658058771, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.413061141967773, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8819358944892883, "num_tokens": 311573638.0, "step": 8169 }, { "epoch": 1.0393079760844677, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.433235168457031, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8581417798995972, "num_tokens": 311611085.0, "step": 8170 }, { "epoch": 1.0394351863630582, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.491415023803711, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8594442009925842, "num_tokens": 311655497.0, "step": 8171 }, { "epoch": 1.0395623966416487, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.436178207397461, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.851613461971283, "num_tokens": 311688167.0, "step": 8172 }, { "epoch": 1.0396896069202393, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.471684455871582, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8650978803634644, "num_tokens": 311724926.0, "step": 8173 }, { "epoch": 1.0398168171988296, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.485499382019043, "learning_rate": 1e-06, "loss": 0.4856, "mean_token_accuracy": 0.8458295464515686, "num_tokens": 311760729.0, "step": 8174 }, { "epoch": 1.03994402747742, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.451836585998535, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8659111261367798, "num_tokens": 311802138.0, "step": 8175 }, { "epoch": 1.0400712377560106, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.449685096740723, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8733548521995544, "num_tokens": 311838510.0, "step": 8176 }, { "epoch": 1.0401984480346012, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.479185104370117, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8582608699798584, "num_tokens": 311870875.0, "step": 8177 }, { "epoch": 1.0403256583131917, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.463240623474121, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8781181573867798, "num_tokens": 311905854.0, "step": 8178 }, { "epoch": 1.0404528685917822, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.520630836486816, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8627134561538696, "num_tokens": 311936381.0, "step": 8179 }, { "epoch": 1.0405800788703727, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.448673248291016, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8593328595161438, "num_tokens": 311972529.0, "step": 8180 }, { "epoch": 1.0407072891489633, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.503033638000488, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8839097023010254, "num_tokens": 312007883.0, "step": 8181 }, { "epoch": 1.0408344994275538, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.484488487243652, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8607574105262756, "num_tokens": 312041000.0, "step": 8182 }, { "epoch": 1.0409617097061443, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.497529029846191, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8700627684593201, "num_tokens": 312076011.0, "step": 8183 }, { "epoch": 1.0410889199847349, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.489904403686523, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8588480949401855, "num_tokens": 312115614.0, "step": 8184 }, { "epoch": 1.0412161302633254, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.46550178527832, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8593884110450745, "num_tokens": 312154111.0, "step": 8185 }, { "epoch": 1.0413433405419157, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.425323486328125, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8597736358642578, "num_tokens": 312191571.0, "step": 8186 }, { "epoch": 1.0414705508205062, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.501110076904297, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8577741980552673, "num_tokens": 312229205.0, "step": 8187 }, { "epoch": 1.0415977610990967, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.48920726776123, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8805629014968872, "num_tokens": 312266604.0, "step": 8188 }, { "epoch": 1.0417249713776873, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.420855522155762, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8563047647476196, "num_tokens": 312305222.0, "step": 8189 }, { "epoch": 1.0418521816562778, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.454400062561035, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8443175554275513, "num_tokens": 312345823.0, "step": 8190 }, { "epoch": 1.0419793919348683, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.490625381469727, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8568165898323059, "num_tokens": 312383569.0, "step": 8191 }, { "epoch": 1.0421066022134589, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.43684196472168, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8687585592269897, "num_tokens": 312423068.0, "step": 8192 }, { "epoch": 1.0422338124920494, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.46523666381836, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8602184653282166, "num_tokens": 312460569.0, "step": 8193 }, { "epoch": 1.04236102277064, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.539323806762695, "learning_rate": 1e-06, "loss": 0.4881, "mean_token_accuracy": 0.8442109823226929, "num_tokens": 312501654.0, "step": 8194 }, { "epoch": 1.0424882330492304, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.415674209594727, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8567817211151123, "num_tokens": 312543630.0, "step": 8195 }, { "epoch": 1.042615443327821, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.489112854003906, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.86268550157547, "num_tokens": 312579428.0, "step": 8196 }, { "epoch": 1.0427426536064115, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.499496459960938, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8797218799591064, "num_tokens": 312621388.0, "step": 8197 }, { "epoch": 1.0428698638850018, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.405948638916016, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8652356863021851, "num_tokens": 312665303.0, "step": 8198 }, { "epoch": 1.0429970741635923, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.504952430725098, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8551232814788818, "num_tokens": 312702354.0, "step": 8199 }, { "epoch": 1.0431242844421829, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.476653099060059, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.870144784450531, "num_tokens": 312737223.0, "step": 8200 }, { "epoch": 1.0432514947207734, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.46182918548584, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8653386831283569, "num_tokens": 312773353.0, "step": 8201 }, { "epoch": 1.043378704999364, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.489319801330566, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8470373153686523, "num_tokens": 312809230.0, "step": 8202 }, { "epoch": 1.0435059152779544, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.424452781677246, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.882664144039154, "num_tokens": 312841278.0, "step": 8203 }, { "epoch": 1.043633125556545, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.481431007385254, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.869077205657959, "num_tokens": 312888469.0, "step": 8204 }, { "epoch": 1.0437603358351355, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.426709175109863, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8716219663619995, "num_tokens": 312925446.0, "step": 8205 }, { "epoch": 1.043887546113726, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.484101295471191, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8639650344848633, "num_tokens": 312965425.0, "step": 8206 }, { "epoch": 1.0440147563923166, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.433343887329102, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8593880534172058, "num_tokens": 313000637.0, "step": 8207 }, { "epoch": 1.044141966670907, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.519644737243652, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8654474020004272, "num_tokens": 313035986.0, "step": 8208 }, { "epoch": 1.0442691769494976, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.439502716064453, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8548524379730225, "num_tokens": 313077027.0, "step": 8209 }, { "epoch": 1.0443963872280881, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.486381530761719, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8758170008659363, "num_tokens": 313121733.0, "step": 8210 }, { "epoch": 1.0445235975066784, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.468477249145508, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8598623871803284, "num_tokens": 313165542.0, "step": 8211 }, { "epoch": 1.044650807785269, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.500409126281738, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8542444705963135, "num_tokens": 313201082.0, "step": 8212 }, { "epoch": 1.0447780180638595, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.412487983703613, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8766976594924927, "num_tokens": 313241765.0, "step": 8213 }, { "epoch": 1.04490522834245, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.523181915283203, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8629368543624878, "num_tokens": 313282471.0, "step": 8214 }, { "epoch": 1.0450324386210406, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.451179504394531, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8607091903686523, "num_tokens": 313331505.0, "step": 8215 }, { "epoch": 1.045159648899631, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.447434425354004, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8653677701950073, "num_tokens": 313369823.0, "step": 8216 }, { "epoch": 1.0452868591782216, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.550041198730469, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8658084869384766, "num_tokens": 313404255.0, "step": 8217 }, { "epoch": 1.0454140694568121, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.4746732711792, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8762270212173462, "num_tokens": 313439368.0, "step": 8218 }, { "epoch": 1.0455412797354027, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.513415336608887, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8568682670593262, "num_tokens": 313479475.0, "step": 8219 }, { "epoch": 1.0456684900139932, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.500391006469727, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8556408882141113, "num_tokens": 313514604.0, "step": 8220 }, { "epoch": 1.0457957002925837, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.470491409301758, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8605799674987793, "num_tokens": 313548907.0, "step": 8221 }, { "epoch": 1.0459229105711743, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.459362030029297, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8650140166282654, "num_tokens": 313586615.0, "step": 8222 }, { "epoch": 1.0460501208497646, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.426528930664062, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8565818667411804, "num_tokens": 313625202.0, "step": 8223 }, { "epoch": 1.046177331128355, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.480560302734375, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8577454090118408, "num_tokens": 313663257.0, "step": 8224 }, { "epoch": 1.0463045414069456, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.461042404174805, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8691627979278564, "num_tokens": 313701430.0, "step": 8225 }, { "epoch": 1.0464317516855361, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.517809867858887, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8765966296195984, "num_tokens": 313741286.0, "step": 8226 }, { "epoch": 1.0465589619641267, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.835978507995605, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.8499158620834351, "num_tokens": 313779115.0, "step": 8227 }, { "epoch": 1.0466861722427172, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.487152099609375, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8697477579116821, "num_tokens": 313822331.0, "step": 8228 }, { "epoch": 1.0468133825213077, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.356589317321777, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.85932856798172, "num_tokens": 313863545.0, "step": 8229 }, { "epoch": 1.0469405927998983, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.566611289978027, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8850250244140625, "num_tokens": 313894956.0, "step": 8230 }, { "epoch": 1.0470678030784888, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.454907417297363, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8727982640266418, "num_tokens": 313927920.0, "step": 8231 }, { "epoch": 1.0471950133570793, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.518342018127441, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.87368243932724, "num_tokens": 313963230.0, "step": 8232 }, { "epoch": 1.0473222236356698, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.540175437927246, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8569872379302979, "num_tokens": 313995951.0, "step": 8233 }, { "epoch": 1.0474494339142604, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.451354026794434, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8655582666397095, "num_tokens": 314040928.0, "step": 8234 }, { "epoch": 1.0475766441928507, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.465713500976562, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8570647239685059, "num_tokens": 314080329.0, "step": 8235 }, { "epoch": 1.0477038544714412, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.52546215057373, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8663240075111389, "num_tokens": 314117107.0, "step": 8236 }, { "epoch": 1.0478310647500317, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.489498138427734, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8626778721809387, "num_tokens": 314148077.0, "step": 8237 }, { "epoch": 1.0479582750286223, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.497982025146484, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8565884232521057, "num_tokens": 314182064.0, "step": 8238 }, { "epoch": 1.0480854853072128, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.444478988647461, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8561223149299622, "num_tokens": 314221667.0, "step": 8239 }, { "epoch": 1.0482126955858033, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.48875904083252, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.862066388130188, "num_tokens": 314255225.0, "step": 8240 }, { "epoch": 1.0483399058643939, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.48547077178955, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.865217387676239, "num_tokens": 314288011.0, "step": 8241 }, { "epoch": 1.0484671161429844, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.467070579528809, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8612601161003113, "num_tokens": 314332461.0, "step": 8242 }, { "epoch": 1.048594326421575, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.472631454467773, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.859893798828125, "num_tokens": 314367967.0, "step": 8243 }, { "epoch": 1.0487215367001654, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.41707992553711, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.874600887298584, "num_tokens": 314404487.0, "step": 8244 }, { "epoch": 1.048848746978756, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.472423553466797, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8627204895019531, "num_tokens": 314444731.0, "step": 8245 }, { "epoch": 1.0489759572573465, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.499166488647461, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8654419779777527, "num_tokens": 314479720.0, "step": 8246 }, { "epoch": 1.0491031675359368, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.51392650604248, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8824872970581055, "num_tokens": 314515613.0, "step": 8247 }, { "epoch": 1.0492303778145273, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.437949180603027, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8632829785346985, "num_tokens": 314556519.0, "step": 8248 }, { "epoch": 1.0493575880931179, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.504044532775879, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8756065964698792, "num_tokens": 314590803.0, "step": 8249 }, { "epoch": 1.0494847983717084, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.48652458190918, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8656682968139648, "num_tokens": 314631347.0, "step": 8250 }, { "epoch": 1.049612008650299, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.43860149383545, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8650319576263428, "num_tokens": 314673739.0, "step": 8251 }, { "epoch": 1.0497392189288894, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.486435890197754, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8646537065505981, "num_tokens": 314718498.0, "step": 8252 }, { "epoch": 1.04986642920748, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.454545974731445, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8707211017608643, "num_tokens": 314758755.0, "step": 8253 }, { "epoch": 1.0499936394860705, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.523072242736816, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.868615448474884, "num_tokens": 314793644.0, "step": 8254 }, { "epoch": 1.050120849764661, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.50253963470459, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8610318899154663, "num_tokens": 314828045.0, "step": 8255 }, { "epoch": 1.0502480600432516, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.532113075256348, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8686972260475159, "num_tokens": 314866088.0, "step": 8256 }, { "epoch": 1.050375270321842, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.529720306396484, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8563966751098633, "num_tokens": 314904757.0, "step": 8257 }, { "epoch": 1.0505024806004326, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.529545783996582, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8571350574493408, "num_tokens": 314941410.0, "step": 8258 }, { "epoch": 1.0506296908790231, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.517364501953125, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8735715746879578, "num_tokens": 314982060.0, "step": 8259 }, { "epoch": 1.0507569011576134, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.485078811645508, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8562424182891846, "num_tokens": 315021812.0, "step": 8260 }, { "epoch": 1.050884111436204, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.545198440551758, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8616504669189453, "num_tokens": 315059520.0, "step": 8261 }, { "epoch": 1.0510113217147945, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.498364448547363, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8596588969230652, "num_tokens": 315098392.0, "step": 8262 }, { "epoch": 1.051138531993385, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.521965026855469, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8538059592247009, "num_tokens": 315137375.0, "step": 8263 }, { "epoch": 1.0512657422719756, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.515127182006836, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8692358136177063, "num_tokens": 315171308.0, "step": 8264 }, { "epoch": 1.051392952550566, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.502939224243164, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8654999732971191, "num_tokens": 315212176.0, "step": 8265 }, { "epoch": 1.0515201628291566, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.522927284240723, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8526128530502319, "num_tokens": 315251321.0, "step": 8266 }, { "epoch": 1.0516473731077471, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.519038200378418, "learning_rate": 1e-06, "loss": 0.4649, "mean_token_accuracy": 0.8516847491264343, "num_tokens": 315288780.0, "step": 8267 }, { "epoch": 1.0517745833863377, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.480393409729004, "learning_rate": 1e-06, "loss": 0.342, "mean_token_accuracy": 0.8869880437850952, "num_tokens": 315325630.0, "step": 8268 }, { "epoch": 1.0519017936649282, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.487873077392578, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8591479063034058, "num_tokens": 315362887.0, "step": 8269 }, { "epoch": 1.0520290039435187, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.482244491577148, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8776081204414368, "num_tokens": 315397552.0, "step": 8270 }, { "epoch": 1.0521562142221093, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.492022514343262, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8752042055130005, "num_tokens": 315436785.0, "step": 8271 }, { "epoch": 1.0522834245006996, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.520586013793945, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8742668628692627, "num_tokens": 315467849.0, "step": 8272 }, { "epoch": 1.05241063477929, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.469292640686035, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8720517754554749, "num_tokens": 315501496.0, "step": 8273 }, { "epoch": 1.0525378450578806, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.462272644042969, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8733218908309937, "num_tokens": 315540878.0, "step": 8274 }, { "epoch": 1.0526650553364711, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.496904373168945, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8795691728591919, "num_tokens": 315578139.0, "step": 8275 }, { "epoch": 1.0527922656150617, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.48390007019043, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8667784929275513, "num_tokens": 315615845.0, "step": 8276 }, { "epoch": 1.0529194758936522, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.551175117492676, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8577134609222412, "num_tokens": 315645407.0, "step": 8277 }, { "epoch": 1.0530466861722427, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.467978477478027, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8727187514305115, "num_tokens": 315687922.0, "step": 8278 }, { "epoch": 1.0531738964508333, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.563385963439941, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.866240382194519, "num_tokens": 315723730.0, "step": 8279 }, { "epoch": 1.0533011067294238, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.482538223266602, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8771731853485107, "num_tokens": 315765662.0, "step": 8280 }, { "epoch": 1.0534283170080143, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.514355659484863, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.841293454170227, "num_tokens": 315808668.0, "step": 8281 }, { "epoch": 1.0535555272866048, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.524615287780762, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8611248135566711, "num_tokens": 315844878.0, "step": 8282 }, { "epoch": 1.0536827375651954, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.475310325622559, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8785954713821411, "num_tokens": 315879472.0, "step": 8283 }, { "epoch": 1.0538099478437857, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.564099311828613, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8601298928260803, "num_tokens": 315915496.0, "step": 8284 }, { "epoch": 1.0539371581223762, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.45544719696045, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8670094013214111, "num_tokens": 315954317.0, "step": 8285 }, { "epoch": 1.0540643684009667, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.538412094116211, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8614619970321655, "num_tokens": 315995518.0, "step": 8286 }, { "epoch": 1.0541915786795573, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.475465774536133, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8788110017776489, "num_tokens": 316040475.0, "step": 8287 }, { "epoch": 1.0543187889581478, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.486065864562988, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8640066385269165, "num_tokens": 316083003.0, "step": 8288 }, { "epoch": 1.0544459992367383, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.559244155883789, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.866059422492981, "num_tokens": 316124687.0, "step": 8289 }, { "epoch": 1.0545732095153288, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.460896492004395, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8533048033714294, "num_tokens": 316170052.0, "step": 8290 }, { "epoch": 1.0547004197939194, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.502233505249023, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8530052304267883, "num_tokens": 316211941.0, "step": 8291 }, { "epoch": 1.05482763007251, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.557400703430176, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8690949082374573, "num_tokens": 316252592.0, "step": 8292 }, { "epoch": 1.0549548403511004, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.460453987121582, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8732274174690247, "num_tokens": 316287250.0, "step": 8293 }, { "epoch": 1.055082050629691, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.500731468200684, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8686639070510864, "num_tokens": 316326308.0, "step": 8294 }, { "epoch": 1.0552092609082815, "ewc_loss": 0.024658203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.500567436218262, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8698406219482422, "num_tokens": 316364519.0, "step": 8295 }, { "epoch": 1.0553364711868718, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.536001205444336, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8694595694541931, "num_tokens": 316401165.0, "step": 8296 }, { "epoch": 1.0554636814654623, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.447580337524414, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8606792688369751, "num_tokens": 316440245.0, "step": 8297 }, { "epoch": 1.0555908917440529, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.483922958374023, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8741647005081177, "num_tokens": 316484969.0, "step": 8298 }, { "epoch": 1.0557181020226434, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.519500732421875, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8836361169815063, "num_tokens": 316520510.0, "step": 8299 }, { "epoch": 1.055845312301234, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.499065399169922, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.849599301815033, "num_tokens": 316561946.0, "step": 8300 }, { "epoch": 1.0559725225798244, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.483759880065918, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.882490336894989, "num_tokens": 316605489.0, "step": 8301 }, { "epoch": 1.056099732858415, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.465184211730957, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8487761616706848, "num_tokens": 316648590.0, "step": 8302 }, { "epoch": 1.0562269431370055, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.501115798950195, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8788503408432007, "num_tokens": 316688169.0, "step": 8303 }, { "epoch": 1.056354153415596, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.471160888671875, "learning_rate": 1e-06, "loss": 0.4905, "mean_token_accuracy": 0.8456541299819946, "num_tokens": 316724031.0, "step": 8304 }, { "epoch": 1.0564813636941865, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.510807037353516, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8653509616851807, "num_tokens": 316769762.0, "step": 8305 }, { "epoch": 1.056608573972777, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.472318649291992, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8731146454811096, "num_tokens": 316814156.0, "step": 8306 }, { "epoch": 1.0567357842513676, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.523787498474121, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8746810555458069, "num_tokens": 316852784.0, "step": 8307 }, { "epoch": 1.0568629945299581, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.527653694152832, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8590770959854126, "num_tokens": 316891992.0, "step": 8308 }, { "epoch": 1.0569902048085484, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.470582962036133, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8804988861083984, "num_tokens": 316932246.0, "step": 8309 }, { "epoch": 1.057117415087139, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.516437530517578, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8579075336456299, "num_tokens": 316972017.0, "step": 8310 }, { "epoch": 1.0572446253657295, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.55569076538086, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8606275320053101, "num_tokens": 317008631.0, "step": 8311 }, { "epoch": 1.05737183564432, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.487604141235352, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8713724613189697, "num_tokens": 317048616.0, "step": 8312 }, { "epoch": 1.0574990459229106, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.463006019592285, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8590961694717407, "num_tokens": 317088675.0, "step": 8313 }, { "epoch": 1.057626256201501, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.544493675231934, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.8867504596710205, "num_tokens": 317120351.0, "step": 8314 }, { "epoch": 1.0577534664800916, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.477680206298828, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8590768575668335, "num_tokens": 317159027.0, "step": 8315 }, { "epoch": 1.0578806767586821, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.574000358581543, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8664113283157349, "num_tokens": 317188892.0, "step": 8316 }, { "epoch": 1.0580078870372727, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.491100311279297, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8709644675254822, "num_tokens": 317229026.0, "step": 8317 }, { "epoch": 1.0581350973158632, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.494126319885254, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8642048239707947, "num_tokens": 317272946.0, "step": 8318 }, { "epoch": 1.0582623075944537, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.47290325164795, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8657961487770081, "num_tokens": 317314625.0, "step": 8319 }, { "epoch": 1.058389517873044, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.522741317749023, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8751747012138367, "num_tokens": 317351450.0, "step": 8320 }, { "epoch": 1.0585167281516346, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.553948402404785, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8502807021141052, "num_tokens": 317391519.0, "step": 8321 }, { "epoch": 1.058643938430225, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.478897094726562, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8594678640365601, "num_tokens": 317430672.0, "step": 8322 }, { "epoch": 1.0587711487088156, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.494583129882812, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8674067854881287, "num_tokens": 317470203.0, "step": 8323 }, { "epoch": 1.0588983589874061, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.5387601852417, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8682464361190796, "num_tokens": 317509594.0, "step": 8324 }, { "epoch": 1.0590255692659967, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.466629028320312, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8630654811859131, "num_tokens": 317549735.0, "step": 8325 }, { "epoch": 1.0591527795445872, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.52782917022705, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8645521402359009, "num_tokens": 317585064.0, "step": 8326 }, { "epoch": 1.0592799898231777, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.508841514587402, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.881366491317749, "num_tokens": 317622289.0, "step": 8327 }, { "epoch": 1.0594072001017683, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.524713516235352, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8538777828216553, "num_tokens": 317662872.0, "step": 8328 }, { "epoch": 1.0595344103803588, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.552885055541992, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8664315938949585, "num_tokens": 317701418.0, "step": 8329 }, { "epoch": 1.0596616206589493, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.502507209777832, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8784990310668945, "num_tokens": 317741416.0, "step": 8330 }, { "epoch": 1.0597888309375398, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.508329391479492, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8707458972930908, "num_tokens": 317778623.0, "step": 8331 }, { "epoch": 1.0599160412161304, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.545930862426758, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8682236075401306, "num_tokens": 317816577.0, "step": 8332 }, { "epoch": 1.0600432514947207, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.509957313537598, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8645433187484741, "num_tokens": 317851532.0, "step": 8333 }, { "epoch": 1.0601704617733112, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.477407455444336, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8666968941688538, "num_tokens": 317886789.0, "step": 8334 }, { "epoch": 1.0602976720519017, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.50479507446289, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8679547905921936, "num_tokens": 317929511.0, "step": 8335 }, { "epoch": 1.0604248823304923, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.464017868041992, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8641829490661621, "num_tokens": 317963767.0, "step": 8336 }, { "epoch": 1.0605520926090828, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.488840103149414, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8829216361045837, "num_tokens": 318000458.0, "step": 8337 }, { "epoch": 1.0606793028876733, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.45260238647461, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8832617998123169, "num_tokens": 318040394.0, "step": 8338 }, { "epoch": 1.0608065131662638, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.49344253540039, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8623390197753906, "num_tokens": 318080859.0, "step": 8339 }, { "epoch": 1.0609337234448544, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.456875801086426, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8828101754188538, "num_tokens": 318115378.0, "step": 8340 }, { "epoch": 1.061060933723445, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.480292320251465, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.858548104763031, "num_tokens": 318150351.0, "step": 8341 }, { "epoch": 1.0611881440020354, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.526122093200684, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8655865788459778, "num_tokens": 318192427.0, "step": 8342 }, { "epoch": 1.061315354280626, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.497129440307617, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8537585735321045, "num_tokens": 318232150.0, "step": 8343 }, { "epoch": 1.0614425645592165, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.510942459106445, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8623871207237244, "num_tokens": 318267742.0, "step": 8344 }, { "epoch": 1.0615697748378068, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.46773910522461, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8708603382110596, "num_tokens": 318307974.0, "step": 8345 }, { "epoch": 1.0616969851163973, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.506987571716309, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8479273319244385, "num_tokens": 318349880.0, "step": 8346 }, { "epoch": 1.0618241953949878, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.566126823425293, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8639697432518005, "num_tokens": 318383869.0, "step": 8347 }, { "epoch": 1.0619514056735784, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.52247428894043, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8817940950393677, "num_tokens": 318419659.0, "step": 8348 }, { "epoch": 1.062078615952169, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 12.013811111450195, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8750789165496826, "num_tokens": 318460795.0, "step": 8349 }, { "epoch": 1.0622058262307594, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.535127639770508, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8567678928375244, "num_tokens": 318499451.0, "step": 8350 }, { "epoch": 1.06233303650935, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.392348289489746, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.869881272315979, "num_tokens": 318533413.0, "step": 8351 }, { "epoch": 1.0624602467879405, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.560905456542969, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8548074960708618, "num_tokens": 318572533.0, "step": 8352 }, { "epoch": 1.062587457066531, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.529923439025879, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8686171770095825, "num_tokens": 318609442.0, "step": 8353 }, { "epoch": 1.0627146673451215, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.540207862854004, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8810489773750305, "num_tokens": 318643475.0, "step": 8354 }, { "epoch": 1.062841877623712, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.490822792053223, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8700332045555115, "num_tokens": 318684223.0, "step": 8355 }, { "epoch": 1.0629690879023026, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.531035423278809, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.867267906665802, "num_tokens": 318717148.0, "step": 8356 }, { "epoch": 1.0630962981808931, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.524066925048828, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8767354488372803, "num_tokens": 318758929.0, "step": 8357 }, { "epoch": 1.0632235084594834, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.504934310913086, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8639487028121948, "num_tokens": 318796438.0, "step": 8358 }, { "epoch": 1.063350718738074, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.541332244873047, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.853781521320343, "num_tokens": 318835732.0, "step": 8359 }, { "epoch": 1.0634779290166645, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.621015548706055, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8648048043251038, "num_tokens": 318872151.0, "step": 8360 }, { "epoch": 1.063605139295255, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.529505729675293, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8554013967514038, "num_tokens": 318911956.0, "step": 8361 }, { "epoch": 1.0637323495738455, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.541826248168945, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8684864044189453, "num_tokens": 318945593.0, "step": 8362 }, { "epoch": 1.063859559852436, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.513105392456055, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.876166820526123, "num_tokens": 318976670.0, "step": 8363 }, { "epoch": 1.0639867701310266, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.55836009979248, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.868322491645813, "num_tokens": 319007798.0, "step": 8364 }, { "epoch": 1.0641139804096171, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.51367473602295, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8560202717781067, "num_tokens": 319044808.0, "step": 8365 }, { "epoch": 1.0642411906882077, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.489063262939453, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8852680921554565, "num_tokens": 319083387.0, "step": 8366 }, { "epoch": 1.0643684009667982, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.519142150878906, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8811079859733582, "num_tokens": 319121897.0, "step": 8367 }, { "epoch": 1.0644956112453887, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.518659591674805, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8546379208564758, "num_tokens": 319157190.0, "step": 8368 }, { "epoch": 1.064622821523979, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.516777038574219, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8770149946212769, "num_tokens": 319188353.0, "step": 8369 }, { "epoch": 1.0647500318025696, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.599845886230469, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8677477836608887, "num_tokens": 319225536.0, "step": 8370 }, { "epoch": 1.06487724208116, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.498106956481934, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8734925985336304, "num_tokens": 319265114.0, "step": 8371 }, { "epoch": 1.0650044523597506, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.515786170959473, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8767011165618896, "num_tokens": 319309280.0, "step": 8372 }, { "epoch": 1.0651316626383411, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.552820205688477, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8781270384788513, "num_tokens": 319346114.0, "step": 8373 }, { "epoch": 1.0652588729169317, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.495623588562012, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8744505643844604, "num_tokens": 319389001.0, "step": 8374 }, { "epoch": 1.0653860831955222, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 11.501015663146973, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8741186857223511, "num_tokens": 319431464.0, "step": 8375 }, { "epoch": 1.0655132934741127, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.528935432434082, "learning_rate": 1e-06, "loss": 0.4898, "mean_token_accuracy": 0.8426723480224609, "num_tokens": 319475638.0, "step": 8376 }, { "epoch": 1.0656405037527032, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.54218578338623, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8765119314193726, "num_tokens": 319515983.0, "step": 8377 }, { "epoch": 1.0657677140312938, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.526538848876953, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8641347289085388, "num_tokens": 319552152.0, "step": 8378 }, { "epoch": 1.0658949243098843, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.561760902404785, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8587058782577515, "num_tokens": 319592963.0, "step": 8379 }, { "epoch": 1.0660221345884748, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.54813003540039, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8532081246376038, "num_tokens": 319635136.0, "step": 8380 }, { "epoch": 1.0661493448670654, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.546627044677734, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8690673112869263, "num_tokens": 319670118.0, "step": 8381 }, { "epoch": 1.0662765551456557, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.552650451660156, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8733793497085571, "num_tokens": 319711057.0, "step": 8382 }, { "epoch": 1.0664037654242462, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.492400169372559, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.864820122718811, "num_tokens": 319746451.0, "step": 8383 }, { "epoch": 1.0665309757028367, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.514528274536133, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.859846830368042, "num_tokens": 319792149.0, "step": 8384 }, { "epoch": 1.0666581859814273, "ewc_loss": 0.02490234375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.494654655456543, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8582686185836792, "num_tokens": 319826508.0, "step": 8385 }, { "epoch": 1.0667853962600178, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.497496604919434, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8684865236282349, "num_tokens": 319867000.0, "step": 8386 }, { "epoch": 1.0669126065386083, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.548762321472168, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8472015261650085, "num_tokens": 319904476.0, "step": 8387 }, { "epoch": 1.0670398168171988, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.501026153564453, "learning_rate": 1e-06, "loss": 0.4886, "mean_token_accuracy": 0.8433847427368164, "num_tokens": 319943279.0, "step": 8388 }, { "epoch": 1.0671670270957894, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.491803169250488, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8629275560379028, "num_tokens": 319986610.0, "step": 8389 }, { "epoch": 1.06729423737438, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.53842544555664, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8600571155548096, "num_tokens": 320022702.0, "step": 8390 }, { "epoch": 1.0674214476529704, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.463569641113281, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.868923008441925, "num_tokens": 320067918.0, "step": 8391 }, { "epoch": 1.067548657931561, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.551589965820312, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.857890248298645, "num_tokens": 320110455.0, "step": 8392 }, { "epoch": 1.0676758682101515, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.474065780639648, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8830596804618835, "num_tokens": 320147675.0, "step": 8393 }, { "epoch": 1.0678030784887418, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.51368522644043, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8535571098327637, "num_tokens": 320183882.0, "step": 8394 }, { "epoch": 1.0679302887673323, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.515609741210938, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8557978868484497, "num_tokens": 320222215.0, "step": 8395 }, { "epoch": 1.0680574990459228, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.539817810058594, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8587100505828857, "num_tokens": 320262418.0, "step": 8396 }, { "epoch": 1.0681847093245134, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.527167320251465, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8593029975891113, "num_tokens": 320297304.0, "step": 8397 }, { "epoch": 1.068311919603104, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.540457725524902, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8683005571365356, "num_tokens": 320330919.0, "step": 8398 }, { "epoch": 1.0684391298816944, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.528091430664062, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.865894615650177, "num_tokens": 320367602.0, "step": 8399 }, { "epoch": 1.068566340160285, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.532206535339355, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8658263683319092, "num_tokens": 320403875.0, "step": 8400 }, { "epoch": 1.0686935504388755, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.534551620483398, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8701106309890747, "num_tokens": 320439149.0, "step": 8401 }, { "epoch": 1.068820760717466, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.506560325622559, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8783531188964844, "num_tokens": 320476786.0, "step": 8402 }, { "epoch": 1.0689479709960565, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.54839038848877, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8507179617881775, "num_tokens": 320516945.0, "step": 8403 }, { "epoch": 1.069075181274647, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.526801109313965, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8807955980300903, "num_tokens": 320552801.0, "step": 8404 }, { "epoch": 1.0692023915532376, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.490227699279785, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8691577315330505, "num_tokens": 320593605.0, "step": 8405 }, { "epoch": 1.0693296018318281, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.542275428771973, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8723623752593994, "num_tokens": 320624476.0, "step": 8406 }, { "epoch": 1.0694568121104184, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.509946823120117, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8729897737503052, "num_tokens": 320667000.0, "step": 8407 }, { "epoch": 1.069584022389009, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.577836036682129, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.870312511920929, "num_tokens": 320699286.0, "step": 8408 }, { "epoch": 1.0697112326675995, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.504469871520996, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8726724982261658, "num_tokens": 320732520.0, "step": 8409 }, { "epoch": 1.06983844294619, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.540844917297363, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8556216955184937, "num_tokens": 320773425.0, "step": 8410 }, { "epoch": 1.0699656532247805, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.52294635772705, "learning_rate": 1e-06, "loss": 0.4967, "mean_token_accuracy": 0.8445302248001099, "num_tokens": 320810859.0, "step": 8411 }, { "epoch": 1.070092863503371, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.51878833770752, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8695468902587891, "num_tokens": 320850266.0, "step": 8412 }, { "epoch": 1.0702200737819616, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.47925853729248, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.865104079246521, "num_tokens": 320892812.0, "step": 8413 }, { "epoch": 1.0703472840605521, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.568769454956055, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8778086304664612, "num_tokens": 320929563.0, "step": 8414 }, { "epoch": 1.0704744943391427, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.504364967346191, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8734989166259766, "num_tokens": 320964010.0, "step": 8415 }, { "epoch": 1.0706017046177332, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.527368545532227, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8640082478523254, "num_tokens": 321006405.0, "step": 8416 }, { "epoch": 1.0707289148963237, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.469029426574707, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8612093925476074, "num_tokens": 321050146.0, "step": 8417 }, { "epoch": 1.070856125174914, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.534989356994629, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8667260408401489, "num_tokens": 321086857.0, "step": 8418 }, { "epoch": 1.0709833354535045, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.489895820617676, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8712906241416931, "num_tokens": 321127310.0, "step": 8419 }, { "epoch": 1.071110545732095, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.544188499450684, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8561329245567322, "num_tokens": 321165881.0, "step": 8420 }, { "epoch": 1.0712377560106856, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.515714645385742, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.859779417514801, "num_tokens": 321204494.0, "step": 8421 }, { "epoch": 1.0713649662892761, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.482660293579102, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8562124967575073, "num_tokens": 321243639.0, "step": 8422 }, { "epoch": 1.0714921765678667, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.544890403747559, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8587451577186584, "num_tokens": 321281399.0, "step": 8423 }, { "epoch": 1.0716193868464572, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.57581615447998, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8533937931060791, "num_tokens": 321316539.0, "step": 8424 }, { "epoch": 1.0717465971250477, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.485394477844238, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8615053296089172, "num_tokens": 321348618.0, "step": 8425 }, { "epoch": 1.0718738074036382, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.500155448913574, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8719215393066406, "num_tokens": 321381674.0, "step": 8426 }, { "epoch": 1.0720010176822288, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.501981735229492, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8821905851364136, "num_tokens": 321420342.0, "step": 8427 }, { "epoch": 1.0721282279608193, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.5272798538208, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.863667368888855, "num_tokens": 321450770.0, "step": 8428 }, { "epoch": 1.0722554382394098, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.53253173828125, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.869269847869873, "num_tokens": 321491450.0, "step": 8429 }, { "epoch": 1.0723826485180004, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.476317405700684, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8605730533599854, "num_tokens": 321531774.0, "step": 8430 }, { "epoch": 1.0725098587965907, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.524484634399414, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8662868738174438, "num_tokens": 321567692.0, "step": 8431 }, { "epoch": 1.0726370690751812, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.530217170715332, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8609592914581299, "num_tokens": 321601168.0, "step": 8432 }, { "epoch": 1.0727642793537717, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.575998306274414, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8477374315261841, "num_tokens": 321636100.0, "step": 8433 }, { "epoch": 1.0728914896323622, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.490206718444824, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.867699146270752, "num_tokens": 321674365.0, "step": 8434 }, { "epoch": 1.0730186999109528, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.531665802001953, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8770656585693359, "num_tokens": 321712267.0, "step": 8435 }, { "epoch": 1.0731459101895433, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.54552173614502, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8635632991790771, "num_tokens": 321749955.0, "step": 8436 }, { "epoch": 1.0732731204681338, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.583505630493164, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8743696808815002, "num_tokens": 321788522.0, "step": 8437 }, { "epoch": 1.0734003307467244, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.552473068237305, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8641396760940552, "num_tokens": 321826314.0, "step": 8438 }, { "epoch": 1.073527541025315, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.566801071166992, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8621910214424133, "num_tokens": 321864851.0, "step": 8439 }, { "epoch": 1.0736547513039054, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.553232192993164, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.87459796667099, "num_tokens": 321901001.0, "step": 8440 }, { "epoch": 1.073781961582496, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.532050132751465, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.864780843257904, "num_tokens": 321941368.0, "step": 8441 }, { "epoch": 1.0739091718610865, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.51252269744873, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8692957758903503, "num_tokens": 321978268.0, "step": 8442 }, { "epoch": 1.0740363821396768, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.539203643798828, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8727455139160156, "num_tokens": 322016969.0, "step": 8443 }, { "epoch": 1.0741635924182673, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.504697799682617, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8592568635940552, "num_tokens": 322059362.0, "step": 8444 }, { "epoch": 1.0742908026968578, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.492566108703613, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8461795449256897, "num_tokens": 322097334.0, "step": 8445 }, { "epoch": 1.0744180129754484, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.580313682556152, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.886172354221344, "num_tokens": 322121988.0, "step": 8446 }, { "epoch": 1.074545223254039, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.534714698791504, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8613917827606201, "num_tokens": 322156602.0, "step": 8447 }, { "epoch": 1.0746724335326294, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.54263973236084, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8626049757003784, "num_tokens": 322194475.0, "step": 8448 }, { "epoch": 1.07479964381122, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.573974609375, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8737534284591675, "num_tokens": 322232158.0, "step": 8449 }, { "epoch": 1.0749268540898105, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.493722915649414, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8743941783905029, "num_tokens": 322272170.0, "step": 8450 }, { "epoch": 1.075054064368401, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.529592514038086, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8712984323501587, "num_tokens": 322305496.0, "step": 8451 }, { "epoch": 1.0751812746469915, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.490861892700195, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8709936738014221, "num_tokens": 322347552.0, "step": 8452 }, { "epoch": 1.075308484925582, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.531635284423828, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8694282174110413, "num_tokens": 322386471.0, "step": 8453 }, { "epoch": 1.0754356952041726, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.522539138793945, "learning_rate": 1e-06, "loss": 0.5195, "mean_token_accuracy": 0.8315542936325073, "num_tokens": 322431019.0, "step": 8454 }, { "epoch": 1.0755629054827631, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.550066947937012, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8844268918037415, "num_tokens": 322471525.0, "step": 8455 }, { "epoch": 1.0756901157613534, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.528603553771973, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8684211373329163, "num_tokens": 322516594.0, "step": 8456 }, { "epoch": 1.075817326039944, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.550256729125977, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8715969324111938, "num_tokens": 322557837.0, "step": 8457 }, { "epoch": 1.0759445363185345, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.510086059570312, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8621556162834167, "num_tokens": 322594165.0, "step": 8458 }, { "epoch": 1.076071746597125, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.581596374511719, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8482334017753601, "num_tokens": 322637083.0, "step": 8459 }, { "epoch": 1.0761989568757155, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.544526100158691, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8648262023925781, "num_tokens": 322673216.0, "step": 8460 }, { "epoch": 1.076326167154306, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.54213809967041, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8689765930175781, "num_tokens": 322709150.0, "step": 8461 }, { "epoch": 1.0764533774328966, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.555663108825684, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8725327849388123, "num_tokens": 322747893.0, "step": 8462 }, { "epoch": 1.0765805877114871, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.584467887878418, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8487070798873901, "num_tokens": 322781667.0, "step": 8463 }, { "epoch": 1.0767077979900777, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.57988452911377, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8655016422271729, "num_tokens": 322818008.0, "step": 8464 }, { "epoch": 1.0768350082686682, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.538317680358887, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8451558351516724, "num_tokens": 322856528.0, "step": 8465 }, { "epoch": 1.0769622185472587, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.552258491516113, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8688145875930786, "num_tokens": 322888798.0, "step": 8466 }, { "epoch": 1.077089428825849, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.587676048278809, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8677181005477905, "num_tokens": 322923301.0, "step": 8467 }, { "epoch": 1.0772166391044395, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.506619453430176, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8771630525588989, "num_tokens": 322964253.0, "step": 8468 }, { "epoch": 1.07734384938303, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.535433769226074, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8643622994422913, "num_tokens": 323006926.0, "step": 8469 }, { "epoch": 1.0774710596616206, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.478594779968262, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8793585896492004, "num_tokens": 323047541.0, "step": 8470 }, { "epoch": 1.0775982699402111, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.629191398620605, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8582049608230591, "num_tokens": 323084272.0, "step": 8471 }, { "epoch": 1.0777254802188017, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.591532707214355, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.877655029296875, "num_tokens": 323116698.0, "step": 8472 }, { "epoch": 1.0778526904973922, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.523490905761719, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8739391565322876, "num_tokens": 323158836.0, "step": 8473 }, { "epoch": 1.0779799007759827, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.547212600708008, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8618794679641724, "num_tokens": 323193777.0, "step": 8474 }, { "epoch": 1.0781071110545732, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.561919212341309, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8667395114898682, "num_tokens": 323232546.0, "step": 8475 }, { "epoch": 1.0782343213331638, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.544772148132324, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8665734529495239, "num_tokens": 323269403.0, "step": 8476 }, { "epoch": 1.0783615316117543, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.61534309387207, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8789029717445374, "num_tokens": 323305982.0, "step": 8477 }, { "epoch": 1.0784887418903448, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.571847915649414, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.861739456653595, "num_tokens": 323341032.0, "step": 8478 }, { "epoch": 1.0786159521689354, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.530282974243164, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8607204556465149, "num_tokens": 323381647.0, "step": 8479 }, { "epoch": 1.0787431624475257, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.551368713378906, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8702108860015869, "num_tokens": 323417185.0, "step": 8480 }, { "epoch": 1.0788703727261162, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.54455280303955, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8653988838195801, "num_tokens": 323454624.0, "step": 8481 }, { "epoch": 1.0789975830047067, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.57970142364502, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8660087585449219, "num_tokens": 323492862.0, "step": 8482 }, { "epoch": 1.0791247932832972, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.559103965759277, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8626086711883545, "num_tokens": 323536012.0, "step": 8483 }, { "epoch": 1.0792520035618878, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.576257705688477, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8704217076301575, "num_tokens": 323573495.0, "step": 8484 }, { "epoch": 1.0793792138404783, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.549718856811523, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8758038282394409, "num_tokens": 323610900.0, "step": 8485 }, { "epoch": 1.0795064241190688, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.571693420410156, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8649374842643738, "num_tokens": 323650983.0, "step": 8486 }, { "epoch": 1.0796336343976594, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.608275413513184, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8582475781440735, "num_tokens": 323690155.0, "step": 8487 }, { "epoch": 1.0797608446762499, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.554424285888672, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8795013427734375, "num_tokens": 323732212.0, "step": 8488 }, { "epoch": 1.0798880549548404, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.572311401367188, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8743126392364502, "num_tokens": 323773400.0, "step": 8489 }, { "epoch": 1.080015265233431, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.538561820983887, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8737437725067139, "num_tokens": 323812753.0, "step": 8490 }, { "epoch": 1.0801424755120215, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.54955768585205, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8686573505401611, "num_tokens": 323855835.0, "step": 8491 }, { "epoch": 1.0802696857906118, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.51486587524414, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8788738250732422, "num_tokens": 323896065.0, "step": 8492 }, { "epoch": 1.0803968960692023, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.537699699401855, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8697928190231323, "num_tokens": 323934351.0, "step": 8493 }, { "epoch": 1.0805241063477928, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.51794147491455, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8566411137580872, "num_tokens": 323968878.0, "step": 8494 }, { "epoch": 1.0806513166263834, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.594267845153809, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8654912710189819, "num_tokens": 324004838.0, "step": 8495 }, { "epoch": 1.080778526904974, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.544885635375977, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8817920088768005, "num_tokens": 324040851.0, "step": 8496 }, { "epoch": 1.0809057371835644, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.544326782226562, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8543232679367065, "num_tokens": 324084675.0, "step": 8497 }, { "epoch": 1.081032947462155, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.55217456817627, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.87192702293396, "num_tokens": 324122136.0, "step": 8498 }, { "epoch": 1.0811601577407455, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.630958557128906, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.864823579788208, "num_tokens": 324153142.0, "step": 8499 }, { "epoch": 1.081287368019336, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.63373851776123, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.8466159105300903, "num_tokens": 324187244.0, "step": 8500 }, { "epoch": 1.0814145782979265, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.593268394470215, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8692643642425537, "num_tokens": 324225644.0, "step": 8501 }, { "epoch": 1.081541788576517, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.56125259399414, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8747742772102356, "num_tokens": 324270394.0, "step": 8502 }, { "epoch": 1.0816689988551076, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.550368309020996, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8780055046081543, "num_tokens": 324311244.0, "step": 8503 }, { "epoch": 1.0817962091336981, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.53918170928955, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.848834753036499, "num_tokens": 324350821.0, "step": 8504 }, { "epoch": 1.0819234194122884, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.569249153137207, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.865343451499939, "num_tokens": 324388255.0, "step": 8505 }, { "epoch": 1.082050629690879, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.571149826049805, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8422322273254395, "num_tokens": 324430068.0, "step": 8506 }, { "epoch": 1.0821778399694695, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.551910400390625, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8741117715835571, "num_tokens": 324465341.0, "step": 8507 }, { "epoch": 1.08230505024806, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.530411720275879, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8745384216308594, "num_tokens": 324502199.0, "step": 8508 }, { "epoch": 1.0824322605266505, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.57163143157959, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8678497672080994, "num_tokens": 324535898.0, "step": 8509 }, { "epoch": 1.082559470805241, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.606846809387207, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.856076717376709, "num_tokens": 324574404.0, "step": 8510 }, { "epoch": 1.0826866810838316, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.525440216064453, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8660876154899597, "num_tokens": 324611743.0, "step": 8511 }, { "epoch": 1.0828138913624221, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.648411750793457, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8573975563049316, "num_tokens": 324653126.0, "step": 8512 }, { "epoch": 1.0829411016410126, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.554229736328125, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8731410503387451, "num_tokens": 324692957.0, "step": 8513 }, { "epoch": 1.0830683119196032, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.607275009155273, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8617851734161377, "num_tokens": 324737214.0, "step": 8514 }, { "epoch": 1.0831955221981937, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.55849838256836, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8771603107452393, "num_tokens": 324773098.0, "step": 8515 }, { "epoch": 1.083322732476784, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.61679458618164, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8561251163482666, "num_tokens": 324815846.0, "step": 8516 }, { "epoch": 1.0834499427553745, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.53994083404541, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8628925681114197, "num_tokens": 324855912.0, "step": 8517 }, { "epoch": 1.083577153033965, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.56153678894043, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8601111769676208, "num_tokens": 324893090.0, "step": 8518 }, { "epoch": 1.0837043633125556, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.575355529785156, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8666937351226807, "num_tokens": 324929779.0, "step": 8519 }, { "epoch": 1.0838315735911461, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.59432601928711, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8716314435005188, "num_tokens": 324961349.0, "step": 8520 }, { "epoch": 1.0839587838697367, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.556126594543457, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8664839863777161, "num_tokens": 324998688.0, "step": 8521 }, { "epoch": 1.0840859941483272, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.602356910705566, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8592023849487305, "num_tokens": 325040601.0, "step": 8522 }, { "epoch": 1.0842132044269177, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.599108695983887, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8712053298950195, "num_tokens": 325081729.0, "step": 8523 }, { "epoch": 1.0843404147055082, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.576349258422852, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8551895022392273, "num_tokens": 325118886.0, "step": 8524 }, { "epoch": 1.0844676249840988, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.631239891052246, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.884271502494812, "num_tokens": 325156417.0, "step": 8525 }, { "epoch": 1.0845948352626893, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 11.535597801208496, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8785552978515625, "num_tokens": 325200902.0, "step": 8526 }, { "epoch": 1.0847220455412798, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.662327766418457, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8687415719032288, "num_tokens": 325233362.0, "step": 8527 }, { "epoch": 1.0848492558198704, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.596415519714355, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8655455708503723, "num_tokens": 325274646.0, "step": 8528 }, { "epoch": 1.0849764660984607, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.587093353271484, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8826699256896973, "num_tokens": 325304735.0, "step": 8529 }, { "epoch": 1.0851036763770512, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.609501838684082, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.870829164981842, "num_tokens": 325339833.0, "step": 8530 }, { "epoch": 1.0852308866556417, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.60167121887207, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.854021430015564, "num_tokens": 325373838.0, "step": 8531 }, { "epoch": 1.0853580969342322, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.57349681854248, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8869365453720093, "num_tokens": 325412913.0, "step": 8532 }, { "epoch": 1.0854853072128228, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.630267143249512, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8715004920959473, "num_tokens": 325451830.0, "step": 8533 }, { "epoch": 1.0856125174914133, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.678122520446777, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8480894565582275, "num_tokens": 325487162.0, "step": 8534 }, { "epoch": 1.0857397277700038, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.552139282226562, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8761169910430908, "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": 11.615151405334473, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8781032562255859, "num_tokens": 325559925.0, "step": 8536 }, { "epoch": 1.0859941483271849, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.56933307647705, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8670535087585449, "num_tokens": 325599980.0, "step": 8537 }, { "epoch": 1.0861213586057754, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.609042167663574, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8510458469390869, "num_tokens": 325631745.0, "step": 8538 }, { "epoch": 1.086248568884366, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.54683780670166, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8633856177330017, "num_tokens": 325673583.0, "step": 8539 }, { "epoch": 1.0863757791629565, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.604750633239746, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8592664003372192, "num_tokens": 325712995.0, "step": 8540 }, { "epoch": 1.0865029894415468, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.589122772216797, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8724313974380493, "num_tokens": 325748259.0, "step": 8541 }, { "epoch": 1.0866301997201373, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.599567413330078, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.872039258480072, "num_tokens": 325787676.0, "step": 8542 }, { "epoch": 1.0867574099987278, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.61806583404541, "learning_rate": 1e-06, "loss": 0.4825, "mean_token_accuracy": 0.8427906036376953, "num_tokens": 325827819.0, "step": 8543 }, { "epoch": 1.0868846202773184, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.599205017089844, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.875309944152832, "num_tokens": 325866328.0, "step": 8544 }, { "epoch": 1.0870118305559089, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.544233322143555, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8615207672119141, "num_tokens": 325906752.0, "step": 8545 }, { "epoch": 1.0871390408344994, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.640799522399902, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8613179922103882, "num_tokens": 325943765.0, "step": 8546 }, { "epoch": 1.08726625111309, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.671793937683105, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8725886344909668, "num_tokens": 325981471.0, "step": 8547 }, { "epoch": 1.0873934613916805, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.579424858093262, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8709287643432617, "num_tokens": 326017253.0, "step": 8548 }, { "epoch": 1.087520671670271, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.65274429321289, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8681223392486572, "num_tokens": 326053484.0, "step": 8549 }, { "epoch": 1.0876478819488615, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.648001670837402, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8719164133071899, "num_tokens": 326084779.0, "step": 8550 }, { "epoch": 1.087775092227452, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.554343223571777, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8651301860809326, "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": 11.662538528442383, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8646773099899292, "num_tokens": 326155991.0, "step": 8552 }, { "epoch": 1.0880295127846331, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.58513069152832, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8627506494522095, "num_tokens": 326195738.0, "step": 8553 }, { "epoch": 1.0881567230632234, "ewc_loss": 0.025146484375, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 11.544075012207031, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8668306469917297, "num_tokens": 326237773.0, "step": 8554 }, { "epoch": 1.088283933341814, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.600489616394043, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8811585307121277, "num_tokens": 326278589.0, "step": 8555 }, { "epoch": 1.0884111436204045, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.506921768188477, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8602250218391418, "num_tokens": 326316862.0, "step": 8556 }, { "epoch": 1.088538353898995, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.61833667755127, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8637708425521851, "num_tokens": 326353892.0, "step": 8557 }, { "epoch": 1.0886655641775855, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.554729461669922, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8709162473678589, "num_tokens": 326389478.0, "step": 8558 }, { "epoch": 1.088792774456176, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.565622329711914, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8620628118515015, "num_tokens": 326428072.0, "step": 8559 }, { "epoch": 1.0889199847347666, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.541583061218262, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8690617084503174, "num_tokens": 326466374.0, "step": 8560 }, { "epoch": 1.0890471950133571, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.62336540222168, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8623275756835938, "num_tokens": 326501081.0, "step": 8561 }, { "epoch": 1.0891744052919476, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.580395698547363, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8757494688034058, "num_tokens": 326532539.0, "step": 8562 }, { "epoch": 1.0893016155705382, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.53404426574707, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8722515106201172, "num_tokens": 326569481.0, "step": 8563 }, { "epoch": 1.0894288258491287, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.635093688964844, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8642210960388184, "num_tokens": 326609746.0, "step": 8564 }, { "epoch": 1.089556036127719, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.611425399780273, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8561457395553589, "num_tokens": 326652379.0, "step": 8565 }, { "epoch": 1.0896832464063095, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.576918601989746, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.86240154504776, "num_tokens": 326687617.0, "step": 8566 }, { "epoch": 1.0898104566849, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.627738952636719, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8584122657775879, "num_tokens": 326724155.0, "step": 8567 }, { "epoch": 1.0899376669634906, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.559453010559082, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8648123741149902, "num_tokens": 326768234.0, "step": 8568 }, { "epoch": 1.0900648772420811, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.650786399841309, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8701381683349609, "num_tokens": 326810423.0, "step": 8569 }, { "epoch": 1.0901920875206716, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.560174942016602, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8653011322021484, "num_tokens": 326851834.0, "step": 8570 }, { "epoch": 1.0903192977992622, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.603287696838379, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8536709547042847, "num_tokens": 326889334.0, "step": 8571 }, { "epoch": 1.0904465080778527, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.57994556427002, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8550719618797302, "num_tokens": 326931884.0, "step": 8572 }, { "epoch": 1.0905737183564432, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.628495216369629, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8738654851913452, "num_tokens": 326969043.0, "step": 8573 }, { "epoch": 1.0907009286350338, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.59238052368164, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8667360544204712, "num_tokens": 326998369.0, "step": 8574 }, { "epoch": 1.0908281389136243, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.595858573913574, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8658262491226196, "num_tokens": 327036909.0, "step": 8575 }, { "epoch": 1.0909553491922148, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.603241920471191, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.859154999256134, "num_tokens": 327076053.0, "step": 8576 }, { "epoch": 1.0910825594708053, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.581130027770996, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8731976747512817, "num_tokens": 327112285.0, "step": 8577 }, { "epoch": 1.0912097697493957, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.576136589050293, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8722315430641174, "num_tokens": 327151591.0, "step": 8578 }, { "epoch": 1.0913369800279862, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.591391563415527, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8505735397338867, "num_tokens": 327193600.0, "step": 8579 }, { "epoch": 1.0914641903065767, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.58879280090332, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8570176959037781, "num_tokens": 327230249.0, "step": 8580 }, { "epoch": 1.0915914005851672, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.596549034118652, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8792552947998047, "num_tokens": 327270043.0, "step": 8581 }, { "epoch": 1.0917186108637578, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.619744300842285, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8690107464790344, "num_tokens": 327309396.0, "step": 8582 }, { "epoch": 1.0918458211423483, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.605093002319336, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8746000528335571, "num_tokens": 327349497.0, "step": 8583 }, { "epoch": 1.0919730314209388, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.662784576416016, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8582894802093506, "num_tokens": 327385648.0, "step": 8584 }, { "epoch": 1.0921002416995294, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.580448150634766, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8669865131378174, "num_tokens": 327426379.0, "step": 8585 }, { "epoch": 1.0922274519781199, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.618621826171875, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8710080981254578, "num_tokens": 327467555.0, "step": 8586 }, { "epoch": 1.0923546622567104, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.601502418518066, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8626062870025635, "num_tokens": 327505303.0, "step": 8587 }, { "epoch": 1.092481872535301, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.634625434875488, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8715862035751343, "num_tokens": 327547854.0, "step": 8588 }, { "epoch": 1.0926090828138915, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.593950271606445, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8677485585212708, "num_tokens": 327584132.0, "step": 8589 }, { "epoch": 1.0927362930924818, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.572244644165039, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8550764918327332, "num_tokens": 327629250.0, "step": 8590 }, { "epoch": 1.0928635033710723, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.655472755432129, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8583129048347473, "num_tokens": 327666988.0, "step": 8591 }, { "epoch": 1.0929907136496628, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.603619575500488, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8687070608139038, "num_tokens": 327705955.0, "step": 8592 }, { "epoch": 1.0931179239282534, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.589621543884277, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8586300611495972, "num_tokens": 327746362.0, "step": 8593 }, { "epoch": 1.0932451342068439, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.614211082458496, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8654266595840454, "num_tokens": 327789034.0, "step": 8594 }, { "epoch": 1.0933723444854344, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.612245559692383, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8561794757843018, "num_tokens": 327824203.0, "step": 8595 }, { "epoch": 1.093499554764025, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.62289810180664, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.867406964302063, "num_tokens": 327860340.0, "step": 8596 }, { "epoch": 1.0936267650426155, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.586627006530762, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8593955039978027, "num_tokens": 327901709.0, "step": 8597 }, { "epoch": 1.093753975321206, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.591049194335938, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8674939870834351, "num_tokens": 327940327.0, "step": 8598 }, { "epoch": 1.0938811855997965, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.61673641204834, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8715214133262634, "num_tokens": 327977966.0, "step": 8599 }, { "epoch": 1.094008395878387, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.59766960144043, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8623058795928955, "num_tokens": 328021143.0, "step": 8600 }, { "epoch": 1.0941356061569776, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.592462539672852, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8715871572494507, "num_tokens": 328060792.0, "step": 8601 }, { "epoch": 1.094262816435568, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.566229820251465, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.866845965385437, "num_tokens": 328099698.0, "step": 8602 }, { "epoch": 1.0943900267141584, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.62076187133789, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8694740533828735, "num_tokens": 328141075.0, "step": 8603 }, { "epoch": 1.094517236992749, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.618093490600586, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8594101667404175, "num_tokens": 328178945.0, "step": 8604 }, { "epoch": 1.0946444472713395, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.62471866607666, "learning_rate": 1e-06, "loss": 0.4911, "mean_token_accuracy": 0.839672327041626, "num_tokens": 328214423.0, "step": 8605 }, { "epoch": 1.09477165754993, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.65868854522705, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8663583993911743, "num_tokens": 328248640.0, "step": 8606 }, { "epoch": 1.0948988678285205, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.56896686553955, "learning_rate": 1e-06, "loss": 0.4692, "mean_token_accuracy": 0.8447924256324768, "num_tokens": 328283016.0, "step": 8607 }, { "epoch": 1.095026078107111, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.639364242553711, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8696260452270508, "num_tokens": 328322377.0, "step": 8608 }, { "epoch": 1.0951532883857016, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.57553482055664, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8672974109649658, "num_tokens": 328359409.0, "step": 8609 }, { "epoch": 1.0952804986642921, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.641083717346191, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8705024719238281, "num_tokens": 328401257.0, "step": 8610 }, { "epoch": 1.0954077089428826, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.601962089538574, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8689560890197754, "num_tokens": 328435886.0, "step": 8611 }, { "epoch": 1.0955349192214732, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.634946823120117, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8556237816810608, "num_tokens": 328472091.0, "step": 8612 }, { "epoch": 1.0956621295000637, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.634766578674316, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8581037521362305, "num_tokens": 328509453.0, "step": 8613 }, { "epoch": 1.095789339778654, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.61415958404541, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8654330372810364, "num_tokens": 328545590.0, "step": 8614 }, { "epoch": 1.0959165500572445, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.574183464050293, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8561015129089355, "num_tokens": 328583093.0, "step": 8615 }, { "epoch": 1.096043760335835, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.631954193115234, "learning_rate": 1e-06, "loss": 0.4707, "mean_token_accuracy": 0.8491250276565552, "num_tokens": 328625312.0, "step": 8616 }, { "epoch": 1.0961709706144256, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.621519088745117, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8585853576660156, "num_tokens": 328655392.0, "step": 8617 }, { "epoch": 1.0962981808930161, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.552011489868164, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8503009080886841, "num_tokens": 328693686.0, "step": 8618 }, { "epoch": 1.0964253911716066, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.620698928833008, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8777799010276794, "num_tokens": 328728051.0, "step": 8619 }, { "epoch": 1.0965526014501972, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.604430198669434, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8558406233787537, "num_tokens": 328768091.0, "step": 8620 }, { "epoch": 1.0966798117287877, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.579931259155273, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8620112538337708, "num_tokens": 328803851.0, "step": 8621 }, { "epoch": 1.0968070220073782, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.593778610229492, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8769652843475342, "num_tokens": 328837429.0, "step": 8622 }, { "epoch": 1.0969342322859688, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.5873441696167, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8647183775901794, "num_tokens": 328874264.0, "step": 8623 }, { "epoch": 1.0970614425645593, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.591621398925781, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8683333992958069, "num_tokens": 328911627.0, "step": 8624 }, { "epoch": 1.0971886528431498, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.604447364807129, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8587895631790161, "num_tokens": 328944471.0, "step": 8625 }, { "epoch": 1.0973158631217403, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.591628074645996, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8731986284255981, "num_tokens": 328983826.0, "step": 8626 }, { "epoch": 1.0974430734003306, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.647428512573242, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8658336400985718, "num_tokens": 329019385.0, "step": 8627 }, { "epoch": 1.0975702836789212, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.608685493469238, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8532475829124451, "num_tokens": 329065179.0, "step": 8628 }, { "epoch": 1.0976974939575117, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.620376586914062, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8722918033599854, "num_tokens": 329101077.0, "step": 8629 }, { "epoch": 1.0978247042361022, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.663996696472168, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8708815574645996, "num_tokens": 329137538.0, "step": 8630 }, { "epoch": 1.0979519145146928, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.613409042358398, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8798896074295044, "num_tokens": 329171952.0, "step": 8631 }, { "epoch": 1.0980791247932833, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.624505996704102, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8741326332092285, "num_tokens": 329214956.0, "step": 8632 }, { "epoch": 1.0982063350718738, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.59370231628418, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8558117151260376, "num_tokens": 329254228.0, "step": 8633 }, { "epoch": 1.0983335453504643, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.683708190917969, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8678210973739624, "num_tokens": 329288337.0, "step": 8634 }, { "epoch": 1.0984607556290549, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.625900268554688, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8595616817474365, "num_tokens": 329329118.0, "step": 8635 }, { "epoch": 1.0985879659076454, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.621210098266602, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8523733615875244, "num_tokens": 329366301.0, "step": 8636 }, { "epoch": 1.098715176186236, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.637798309326172, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8541257381439209, "num_tokens": 329401612.0, "step": 8637 }, { "epoch": 1.0988423864648265, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.680144309997559, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.868018388748169, "num_tokens": 329440403.0, "step": 8638 }, { "epoch": 1.0989695967434168, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.581583976745605, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8752825260162354, "num_tokens": 329478699.0, "step": 8639 }, { "epoch": 1.0990968070220073, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.70077896118164, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8640830516815186, "num_tokens": 329517860.0, "step": 8640 }, { "epoch": 1.0992240173005978, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.620037078857422, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8713743090629578, "num_tokens": 329555385.0, "step": 8641 }, { "epoch": 1.0993512275791884, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.60147476196289, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8625016212463379, "num_tokens": 329594775.0, "step": 8642 }, { "epoch": 1.0994784378577789, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.678189277648926, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8696271181106567, "num_tokens": 329629872.0, "step": 8643 }, { "epoch": 1.0996056481363694, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.625802993774414, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.860041618347168, "num_tokens": 329665263.0, "step": 8644 }, { "epoch": 1.09973285841496, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.604449272155762, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8558422327041626, "num_tokens": 329705853.0, "step": 8645 }, { "epoch": 1.0998600686935505, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.635065078735352, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8726223111152649, "num_tokens": 329741690.0, "step": 8646 }, { "epoch": 1.099987278972141, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.61103630065918, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8581219911575317, "num_tokens": 329781548.0, "step": 8647 }, { "epoch": 1.1001144892507315, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.626507759094238, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8728631734848022, "num_tokens": 329822304.0, "step": 8648 }, { "epoch": 1.100241699529322, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.617753982543945, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8671687841415405, "num_tokens": 329856065.0, "step": 8649 }, { "epoch": 1.1003689098079126, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.627277374267578, "learning_rate": 1e-06, "loss": 0.4861, "mean_token_accuracy": 0.850744903087616, "num_tokens": 329895002.0, "step": 8650 }, { "epoch": 1.100496120086503, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.582295417785645, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.860037088394165, "num_tokens": 329938810.0, "step": 8651 }, { "epoch": 1.1006233303650934, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.590316772460938, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8792918920516968, "num_tokens": 329976513.0, "step": 8652 }, { "epoch": 1.100750540643684, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.60022258758545, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8735278844833374, "num_tokens": 330014013.0, "step": 8653 }, { "epoch": 1.1008777509222745, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.63253402709961, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8612397313117981, "num_tokens": 330052577.0, "step": 8654 }, { "epoch": 1.101004961200865, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.578845977783203, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8630305528640747, "num_tokens": 330091385.0, "step": 8655 }, { "epoch": 1.1011321714794555, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.621267318725586, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8742448091506958, "num_tokens": 330130285.0, "step": 8656 }, { "epoch": 1.101259381758046, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.653152465820312, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8518917560577393, "num_tokens": 330166823.0, "step": 8657 }, { "epoch": 1.1013865920366366, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.60265827178955, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8670940399169922, "num_tokens": 330200655.0, "step": 8658 }, { "epoch": 1.101513802315227, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.600892066955566, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8642597794532776, "num_tokens": 330243656.0, "step": 8659 }, { "epoch": 1.1016410125938176, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.652085304260254, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8569320440292358, "num_tokens": 330287200.0, "step": 8660 }, { "epoch": 1.1017682228724082, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.59847354888916, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8685867786407471, "num_tokens": 330324794.0, "step": 8661 }, { "epoch": 1.1018954331509987, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.633919715881348, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8677676916122437, "num_tokens": 330365297.0, "step": 8662 }, { "epoch": 1.102022643429589, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.581648826599121, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8827382922172546, "num_tokens": 330408046.0, "step": 8663 }, { "epoch": 1.1021498537081795, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.633905410766602, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8619447350502014, "num_tokens": 330447815.0, "step": 8664 }, { "epoch": 1.10227706398677, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.62513256072998, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8548553586006165, "num_tokens": 330487796.0, "step": 8665 }, { "epoch": 1.1024042742653606, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.62497329711914, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8768378496170044, "num_tokens": 330531155.0, "step": 8666 }, { "epoch": 1.1025314845439511, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.617799758911133, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8751533031463623, "num_tokens": 330568611.0, "step": 8667 }, { "epoch": 1.1026586948225416, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.670967102050781, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8605005145072937, "num_tokens": 330608085.0, "step": 8668 }, { "epoch": 1.1027859051011322, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.667875289916992, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.848874568939209, "num_tokens": 330644776.0, "step": 8669 }, { "epoch": 1.1029131153797227, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.650184631347656, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8563830256462097, "num_tokens": 330686401.0, "step": 8670 }, { "epoch": 1.1030403256583132, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.659697532653809, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8701746463775635, "num_tokens": 330723115.0, "step": 8671 }, { "epoch": 1.1031675359369038, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.58713436126709, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8807957172393799, "num_tokens": 330761170.0, "step": 8672 }, { "epoch": 1.1032947462154943, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.668120384216309, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8664292097091675, "num_tokens": 330790230.0, "step": 8673 }, { "epoch": 1.1034219564940848, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.671093940734863, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8763126134872437, "num_tokens": 330826178.0, "step": 8674 }, { "epoch": 1.1035491667726753, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.63752269744873, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8728073835372925, "num_tokens": 330866341.0, "step": 8675 }, { "epoch": 1.1036763770512656, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.62801456451416, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8526285886764526, "num_tokens": 330903955.0, "step": 8676 }, { "epoch": 1.1038035873298562, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.598098754882812, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8614768385887146, "num_tokens": 330943024.0, "step": 8677 }, { "epoch": 1.1039307976084467, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.600220680236816, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8596796989440918, "num_tokens": 330981004.0, "step": 8678 }, { "epoch": 1.1040580078870372, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.6144437789917, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8650997877120972, "num_tokens": 331021532.0, "step": 8679 }, { "epoch": 1.1041852181656278, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.629085540771484, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8732490539550781, "num_tokens": 331058010.0, "step": 8680 }, { "epoch": 1.1043124284442183, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.660359382629395, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8736798763275146, "num_tokens": 331090713.0, "step": 8681 }, { "epoch": 1.1044396387228088, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.657646179199219, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8716196417808533, "num_tokens": 331127532.0, "step": 8682 }, { "epoch": 1.1045668490013993, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.658262252807617, "learning_rate": 1e-06, "loss": 0.4885, "mean_token_accuracy": 0.8460209965705872, "num_tokens": 331172088.0, "step": 8683 }, { "epoch": 1.1046940592799899, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.679356575012207, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8616528511047363, "num_tokens": 331213419.0, "step": 8684 }, { "epoch": 1.1048212695585804, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.617807388305664, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8652804493904114, "num_tokens": 331250102.0, "step": 8685 }, { "epoch": 1.104948479837171, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.691033363342285, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8674301505088806, "num_tokens": 331284597.0, "step": 8686 }, { "epoch": 1.1050756901157615, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.58246898651123, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8584292531013489, "num_tokens": 331330688.0, "step": 8687 }, { "epoch": 1.1052029003943518, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.64549732208252, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8683291673660278, "num_tokens": 331368777.0, "step": 8688 }, { "epoch": 1.1053301106729423, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.581470489501953, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8703693151473999, "num_tokens": 331414369.0, "step": 8689 }, { "epoch": 1.1054573209515328, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.627800941467285, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8722364902496338, "num_tokens": 331452788.0, "step": 8690 }, { "epoch": 1.1055845312301233, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.66059684753418, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8549758791923523, "num_tokens": 331489575.0, "step": 8691 }, { "epoch": 1.1057117415087139, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.629914283752441, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8828992247581482, "num_tokens": 331529129.0, "step": 8692 }, { "epoch": 1.1058389517873044, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.64553165435791, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8615334630012512, "num_tokens": 331564989.0, "step": 8693 }, { "epoch": 1.105966162065895, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.661873817443848, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8664497137069702, "num_tokens": 331602561.0, "step": 8694 }, { "epoch": 1.1060933723444855, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.614869117736816, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8669060468673706, "num_tokens": 331646902.0, "step": 8695 }, { "epoch": 1.106220582623076, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.693203926086426, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8663809299468994, "num_tokens": 331682347.0, "step": 8696 }, { "epoch": 1.1063477929016665, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.613692283630371, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8576045036315918, "num_tokens": 331725813.0, "step": 8697 }, { "epoch": 1.106475003180257, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.653829574584961, "learning_rate": 1e-06, "loss": 0.4803, "mean_token_accuracy": 0.8459028601646423, "num_tokens": 331767856.0, "step": 8698 }, { "epoch": 1.1066022134588476, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.653571128845215, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.874444842338562, "num_tokens": 331804238.0, "step": 8699 }, { "epoch": 1.106729423737438, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.60200023651123, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8665467500686646, "num_tokens": 331839963.0, "step": 8700 }, { "epoch": 1.1068566340160284, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.65286922454834, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8699012398719788, "num_tokens": 331881563.0, "step": 8701 }, { "epoch": 1.106983844294619, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.638928413391113, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8628264665603638, "num_tokens": 331925540.0, "step": 8702 }, { "epoch": 1.1071110545732095, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.610126495361328, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8734090328216553, "num_tokens": 331963171.0, "step": 8703 }, { "epoch": 1.1072382648518, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.641914367675781, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8855713605880737, "num_tokens": 331998310.0, "step": 8704 }, { "epoch": 1.1073654751303905, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.621644973754883, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8715062141418457, "num_tokens": 332034181.0, "step": 8705 }, { "epoch": 1.107492685408981, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.639962196350098, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.875276505947113, "num_tokens": 332075368.0, "step": 8706 }, { "epoch": 1.1076198956875716, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.605260848999023, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8768934607505798, "num_tokens": 332113281.0, "step": 8707 }, { "epoch": 1.107747105966162, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.69115161895752, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8582230806350708, "num_tokens": 332154206.0, "step": 8708 }, { "epoch": 1.1078743162447526, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.618560791015625, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8505807518959045, "num_tokens": 332195409.0, "step": 8709 }, { "epoch": 1.1080015265233432, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.618263244628906, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8749959468841553, "num_tokens": 332233836.0, "step": 8710 }, { "epoch": 1.1081287368019337, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.694795608520508, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8705031871795654, "num_tokens": 332274167.0, "step": 8711 }, { "epoch": 1.108255947080524, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.619166374206543, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8633337616920471, "num_tokens": 332315856.0, "step": 8712 }, { "epoch": 1.1083831573591145, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.695220947265625, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8652157187461853, "num_tokens": 332366365.0, "step": 8713 }, { "epoch": 1.108510367637705, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.631537437438965, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8631823062896729, "num_tokens": 332400411.0, "step": 8714 }, { "epoch": 1.1086375779162956, "ewc_loss": 0.025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 11.618725776672363, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8689190149307251, "num_tokens": 332441959.0, "step": 8715 }, { "epoch": 1.108764788194886, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.647812843322754, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8530310392379761, "num_tokens": 332479545.0, "step": 8716 }, { "epoch": 1.1088919984734766, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.661981582641602, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8768146634101868, "num_tokens": 332516240.0, "step": 8717 }, { "epoch": 1.1090192087520672, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.667919158935547, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8715763092041016, "num_tokens": 332553488.0, "step": 8718 }, { "epoch": 1.1091464190306577, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.70273208618164, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8572771549224854, "num_tokens": 332588646.0, "step": 8719 }, { "epoch": 1.1092736293092482, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.59938907623291, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8562963604927063, "num_tokens": 332629103.0, "step": 8720 }, { "epoch": 1.1094008395878387, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.665740013122559, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.871650218963623, "num_tokens": 332664820.0, "step": 8721 }, { "epoch": 1.1095280498664293, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.616781234741211, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8675122261047363, "num_tokens": 332699772.0, "step": 8722 }, { "epoch": 1.1096552601450198, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.66482162475586, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8746027946472168, "num_tokens": 332733478.0, "step": 8723 }, { "epoch": 1.1097824704236103, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.607346534729004, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8615368604660034, "num_tokens": 332774733.0, "step": 8724 }, { "epoch": 1.1099096807022006, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.690985679626465, "learning_rate": 1e-06, "loss": 0.4837, "mean_token_accuracy": 0.8432943820953369, "num_tokens": 332812970.0, "step": 8725 }, { "epoch": 1.1100368909807912, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.627772331237793, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8623113632202148, "num_tokens": 332854974.0, "step": 8726 }, { "epoch": 1.1101641012593817, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.68153190612793, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8523596525192261, "num_tokens": 332886462.0, "step": 8727 }, { "epoch": 1.1102913115379722, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.641937255859375, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8622065782546997, "num_tokens": 332928614.0, "step": 8728 }, { "epoch": 1.1104185218165628, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.673615455627441, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8664968609809875, "num_tokens": 332962447.0, "step": 8729 }, { "epoch": 1.1105457320951533, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.703527450561523, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8574517965316772, "num_tokens": 332991395.0, "step": 8730 }, { "epoch": 1.1106729423737438, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.633886337280273, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8585686087608337, "num_tokens": 333027852.0, "step": 8731 }, { "epoch": 1.1108001526523343, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.644218444824219, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8685651421546936, "num_tokens": 333063337.0, "step": 8732 }, { "epoch": 1.1109273629309249, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.685877799987793, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8570802807807922, "num_tokens": 333094632.0, "step": 8733 }, { "epoch": 1.1110545732095154, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.652148246765137, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8795796036720276, "num_tokens": 333134071.0, "step": 8734 }, { "epoch": 1.111181783488106, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.580157279968262, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8715782165527344, "num_tokens": 333171288.0, "step": 8735 }, { "epoch": 1.1113089937666965, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.715496063232422, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8731460571289062, "num_tokens": 333212111.0, "step": 8736 }, { "epoch": 1.1114362040452868, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.6423978805542, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8718255758285522, "num_tokens": 333247557.0, "step": 8737 }, { "epoch": 1.1115634143238773, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.628846168518066, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.864991307258606, "num_tokens": 333278705.0, "step": 8738 }, { "epoch": 1.1116906246024678, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.71302318572998, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8678981065750122, "num_tokens": 333316493.0, "step": 8739 }, { "epoch": 1.1118178348810583, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.618969917297363, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8684486150741577, "num_tokens": 333360384.0, "step": 8740 }, { "epoch": 1.1119450451596489, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.661749839782715, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8669565916061401, "num_tokens": 333395568.0, "step": 8741 }, { "epoch": 1.1120722554382394, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.769048690795898, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8703892230987549, "num_tokens": 333429267.0, "step": 8742 }, { "epoch": 1.11219946571683, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.60221004486084, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8725650906562805, "num_tokens": 333469908.0, "step": 8743 }, { "epoch": 1.1123266759954205, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.654857635498047, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8701694011688232, "num_tokens": 333507865.0, "step": 8744 }, { "epoch": 1.112453886274011, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.653861045837402, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8772808313369751, "num_tokens": 333544011.0, "step": 8745 }, { "epoch": 1.1125810965526015, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.598007202148438, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8676589727401733, "num_tokens": 333581949.0, "step": 8746 }, { "epoch": 1.112708306831192, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.657684326171875, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8739255666732788, "num_tokens": 333615994.0, "step": 8747 }, { "epoch": 1.1128355171097826, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.617256164550781, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8820667266845703, "num_tokens": 333654454.0, "step": 8748 }, { "epoch": 1.112962727388373, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.622020721435547, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8593810796737671, "num_tokens": 333700206.0, "step": 8749 }, { "epoch": 1.1130899376669634, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.678522109985352, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8696423172950745, "num_tokens": 333737538.0, "step": 8750 }, { "epoch": 1.113217147945554, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.654631614685059, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8681437969207764, "num_tokens": 333771237.0, "step": 8751 }, { "epoch": 1.1133443582241445, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.714136123657227, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.862980306148529, "num_tokens": 333806522.0, "step": 8752 }, { "epoch": 1.113471568502735, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.687854766845703, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8662651777267456, "num_tokens": 333846656.0, "step": 8753 }, { "epoch": 1.1135987787813255, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.684706687927246, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8582383394241333, "num_tokens": 333886294.0, "step": 8754 }, { "epoch": 1.113725989059916, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.672202110290527, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8630893230438232, "num_tokens": 333923868.0, "step": 8755 }, { "epoch": 1.1138531993385066, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.706745147705078, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8702914714813232, "num_tokens": 333957187.0, "step": 8756 }, { "epoch": 1.113980409617097, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.628201484680176, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8799634575843811, "num_tokens": 333999061.0, "step": 8757 }, { "epoch": 1.1141076198956876, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.603325843811035, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8606322407722473, "num_tokens": 334043595.0, "step": 8758 }, { "epoch": 1.1142348301742782, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.6524076461792, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8707977533340454, "num_tokens": 334078487.0, "step": 8759 }, { "epoch": 1.1143620404528687, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.624730110168457, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8747615814208984, "num_tokens": 334119610.0, "step": 8760 }, { "epoch": 1.114489250731459, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.740889549255371, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8626863360404968, "num_tokens": 334160149.0, "step": 8761 }, { "epoch": 1.1146164610100495, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.61246109008789, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8676385283470154, "num_tokens": 334198667.0, "step": 8762 }, { "epoch": 1.11474367128864, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.690390586853027, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.868120551109314, "num_tokens": 334233838.0, "step": 8763 }, { "epoch": 1.1148708815672306, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.667728424072266, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8584122061729431, "num_tokens": 334275703.0, "step": 8764 }, { "epoch": 1.114998091845821, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.643839836120605, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8780146837234497, "num_tokens": 334311119.0, "step": 8765 }, { "epoch": 1.1151253021244116, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.642959594726562, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8595757484436035, "num_tokens": 334348298.0, "step": 8766 }, { "epoch": 1.1152525124030022, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.674920082092285, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8533873558044434, "num_tokens": 334386986.0, "step": 8767 }, { "epoch": 1.1153797226815927, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.682353973388672, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8711417317390442, "num_tokens": 334423263.0, "step": 8768 }, { "epoch": 1.1155069329601832, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.614120483398438, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8822880983352661, "num_tokens": 334460363.0, "step": 8769 }, { "epoch": 1.1156341432387737, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.667448997497559, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8706797361373901, "num_tokens": 334501883.0, "step": 8770 }, { "epoch": 1.1157613535173643, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.678735733032227, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8663971424102783, "num_tokens": 334536242.0, "step": 8771 }, { "epoch": 1.1158885637959548, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.674230575561523, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8779085278511047, "num_tokens": 334569685.0, "step": 8772 }, { "epoch": 1.1160157740745453, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.652775764465332, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8686684370040894, "num_tokens": 334604946.0, "step": 8773 }, { "epoch": 1.1161429843531356, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.64078426361084, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8732104301452637, "num_tokens": 334645495.0, "step": 8774 }, { "epoch": 1.1162701946317262, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.695158004760742, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8615332841873169, "num_tokens": 334684239.0, "step": 8775 }, { "epoch": 1.1163974049103167, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.733241081237793, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8722949624061584, "num_tokens": 334713799.0, "step": 8776 }, { "epoch": 1.1165246151889072, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.649186134338379, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8873425722122192, "num_tokens": 334753374.0, "step": 8777 }, { "epoch": 1.1166518254674977, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.643062591552734, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8683643341064453, "num_tokens": 334792867.0, "step": 8778 }, { "epoch": 1.1167790357460883, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.623144149780273, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8634587526321411, "num_tokens": 334832664.0, "step": 8779 }, { "epoch": 1.1169062460246788, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.730555534362793, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8598810434341431, "num_tokens": 334876763.0, "step": 8780 }, { "epoch": 1.1170334563032693, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.69880485534668, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8682832717895508, "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": 11.70472240447998, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8653733730316162, "num_tokens": 334941939.0, "step": 8782 }, { "epoch": 1.1172878768604504, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.635565757751465, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8752082586288452, "num_tokens": 334981338.0, "step": 8783 }, { "epoch": 1.117415087139041, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.678739547729492, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8667054176330566, "num_tokens": 335020660.0, "step": 8784 }, { "epoch": 1.1175422974176314, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.631771087646484, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8696284890174866, "num_tokens": 335057377.0, "step": 8785 }, { "epoch": 1.1176695076962218, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.68786907196045, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8570203185081482, "num_tokens": 335092912.0, "step": 8786 }, { "epoch": 1.1177967179748123, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.646429061889648, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.881550669670105, "num_tokens": 335129157.0, "step": 8787 }, { "epoch": 1.1179239282534028, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.670825958251953, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8681620955467224, "num_tokens": 335168065.0, "step": 8788 }, { "epoch": 1.1180511385319933, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.598420143127441, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8659379482269287, "num_tokens": 335218876.0, "step": 8789 }, { "epoch": 1.1181783488105839, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.638839721679688, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8769218921661377, "num_tokens": 335256300.0, "step": 8790 }, { "epoch": 1.1183055590891744, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.659507751464844, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8771902322769165, "num_tokens": 335291035.0, "step": 8791 }, { "epoch": 1.118432769367765, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.672097206115723, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8779593706130981, "num_tokens": 335330092.0, "step": 8792 }, { "epoch": 1.1185599796463555, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.67166805267334, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8534042835235596, "num_tokens": 335367146.0, "step": 8793 }, { "epoch": 1.118687189924946, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.625313758850098, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.866569995880127, "num_tokens": 335405069.0, "step": 8794 }, { "epoch": 1.1188144002035365, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.631124496459961, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8572245240211487, "num_tokens": 335443779.0, "step": 8795 }, { "epoch": 1.118941610482127, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.620035171508789, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.859681248664856, "num_tokens": 335485978.0, "step": 8796 }, { "epoch": 1.1190688207607176, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.618008613586426, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8712823987007141, "num_tokens": 335530375.0, "step": 8797 }, { "epoch": 1.119196031039308, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.665685653686523, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8725512027740479, "num_tokens": 335569109.0, "step": 8798 }, { "epoch": 1.1193232413178984, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.698493957519531, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8528018593788147, "num_tokens": 335599283.0, "step": 8799 }, { "epoch": 1.119450451596489, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.672181129455566, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8738923072814941, "num_tokens": 335634792.0, "step": 8800 }, { "epoch": 1.1195776618750795, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.672279357910156, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8804967999458313, "num_tokens": 335672869.0, "step": 8801 }, { "epoch": 1.11970487215367, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.68071174621582, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8721673488616943, "num_tokens": 335714996.0, "step": 8802 }, { "epoch": 1.1198320824322605, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.664313316345215, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8583805561065674, "num_tokens": 335756884.0, "step": 8803 }, { "epoch": 1.119959292710851, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.693387031555176, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.876573920249939, "num_tokens": 335793935.0, "step": 8804 }, { "epoch": 1.1200865029894416, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.64179515838623, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8788892030715942, "num_tokens": 335834169.0, "step": 8805 }, { "epoch": 1.120213713268032, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.729727745056152, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8852593898773193, "num_tokens": 335872302.0, "step": 8806 }, { "epoch": 1.1203409235466226, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.687272071838379, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8669242262840271, "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": 11.704452514648438, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8780792951583862, "num_tokens": 335946268.0, "step": 8808 }, { "epoch": 1.1205953441038037, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.67127799987793, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8678719997406006, "num_tokens": 335985364.0, "step": 8809 }, { "epoch": 1.120722554382394, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.705846786499023, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8707315921783447, "num_tokens": 336025300.0, "step": 8810 }, { "epoch": 1.1208497646609845, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.717385292053223, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.871662974357605, "num_tokens": 336062105.0, "step": 8811 }, { "epoch": 1.120976974939575, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.692237854003906, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8563250303268433, "num_tokens": 336102944.0, "step": 8812 }, { "epoch": 1.1211041852181656, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.617707252502441, "learning_rate": 1e-06, "loss": 0.33, "mean_token_accuracy": 0.8926962614059448, "num_tokens": 336142964.0, "step": 8813 }, { "epoch": 1.121231395496756, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.660943031311035, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8701172471046448, "num_tokens": 336186962.0, "step": 8814 }, { "epoch": 1.1213586057753466, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.65971565246582, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8762121200561523, "num_tokens": 336220775.0, "step": 8815 }, { "epoch": 1.1214858160539372, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.685569763183594, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8641829490661621, "num_tokens": 336263067.0, "step": 8816 }, { "epoch": 1.1216130263325277, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.629018783569336, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8710947036743164, "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": 11.635065078735352, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8807599544525146, "num_tokens": 336346965.0, "step": 8818 }, { "epoch": 1.1218674468897087, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.74885082244873, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8697153925895691, "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": 11.685859680175781, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8718374371528625, "num_tokens": 336414020.0, "step": 8820 }, { "epoch": 1.1221218674468898, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.615596771240234, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8656768798828125, "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": 11.625785827636719, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8787257671356201, "num_tokens": 336484593.0, "step": 8822 }, { "epoch": 1.1223762880040706, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.716739654541016, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.851550817489624, "num_tokens": 336518964.0, "step": 8823 }, { "epoch": 1.1225034982826612, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.644733428955078, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8547030687332153, "num_tokens": 336556320.0, "step": 8824 }, { "epoch": 1.1226307085612517, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 11.677847862243652, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8582425117492676, "num_tokens": 336595473.0, "step": 8825 }, { "epoch": 1.1227579188398422, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.662557601928711, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8675843477249146, "num_tokens": 336631793.0, "step": 8826 }, { "epoch": 1.1228851291184327, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.710820198059082, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8537169694900513, "num_tokens": 336671230.0, "step": 8827 }, { "epoch": 1.1230123393970233, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.65073299407959, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8687356114387512, "num_tokens": 336712015.0, "step": 8828 }, { "epoch": 1.1231395496756138, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.768980026245117, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8524616360664368, "num_tokens": 336741206.0, "step": 8829 }, { "epoch": 1.1232667599542043, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.664769172668457, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8594635725021362, "num_tokens": 336778838.0, "step": 8830 }, { "epoch": 1.1233939702327949, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.679317474365234, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8738741874694824, "num_tokens": 336815471.0, "step": 8831 }, { "epoch": 1.1235211805113854, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.69102954864502, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8696083426475525, "num_tokens": 336850870.0, "step": 8832 }, { "epoch": 1.123648390789976, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.679780006408691, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8717394471168518, "num_tokens": 336891779.0, "step": 8833 }, { "epoch": 1.1237756010685664, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.626677513122559, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8782262206077576, "num_tokens": 336931171.0, "step": 8834 }, { "epoch": 1.1239028113471567, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.694539070129395, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8772518038749695, "num_tokens": 336963080.0, "step": 8835 }, { "epoch": 1.1240300216257473, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.700658798217773, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8539048433303833, "num_tokens": 337000817.0, "step": 8836 }, { "epoch": 1.1241572319043378, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.692099571228027, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8618969321250916, "num_tokens": 337037249.0, "step": 8837 }, { "epoch": 1.1242844421829283, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.694360733032227, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8758041262626648, "num_tokens": 337077752.0, "step": 8838 }, { "epoch": 1.1244116524615189, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.642595291137695, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8651796579360962, "num_tokens": 337119649.0, "step": 8839 }, { "epoch": 1.1245388627401094, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.709108352661133, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.851355254650116, "num_tokens": 337154376.0, "step": 8840 }, { "epoch": 1.1246660730187, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.690010070800781, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8762168884277344, "num_tokens": 337193377.0, "step": 8841 }, { "epoch": 1.1247932832972904, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.705241203308105, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8577683568000793, "num_tokens": 337233176.0, "step": 8842 }, { "epoch": 1.124920493575881, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.718873023986816, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8533389568328857, "num_tokens": 337262995.0, "step": 8843 }, { "epoch": 1.1250477038544715, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.650919914245605, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8667178750038147, "num_tokens": 337302466.0, "step": 8844 }, { "epoch": 1.125174914133062, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.661495208740234, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8780933618545532, "num_tokens": 337341197.0, "step": 8845 }, { "epoch": 1.1253021244116526, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.748729705810547, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8685898184776306, "num_tokens": 337375602.0, "step": 8846 }, { "epoch": 1.125429334690243, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.672191619873047, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8590399622917175, "num_tokens": 337406685.0, "step": 8847 }, { "epoch": 1.1255565449688334, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.610246658325195, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8624253869056702, "num_tokens": 337447364.0, "step": 8848 }, { "epoch": 1.125683755247424, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.667923927307129, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8663347363471985, "num_tokens": 337487840.0, "step": 8849 }, { "epoch": 1.1258109655260145, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.62894058227539, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8722569346427917, "num_tokens": 337525117.0, "step": 8850 }, { "epoch": 1.125938175804605, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.738973617553711, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8613036870956421, "num_tokens": 337567024.0, "step": 8851 }, { "epoch": 1.1260653860831955, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.655975341796875, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8654025197029114, "num_tokens": 337600117.0, "step": 8852 }, { "epoch": 1.126192596361786, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.695093154907227, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8528472781181335, "num_tokens": 337637188.0, "step": 8853 }, { "epoch": 1.1263198066403766, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.704297065734863, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8619457483291626, "num_tokens": 337676492.0, "step": 8854 }, { "epoch": 1.126447016918967, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.699573516845703, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8803095817565918, "num_tokens": 337713468.0, "step": 8855 }, { "epoch": 1.1265742271975576, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.75599193572998, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.854327917098999, "num_tokens": 337750944.0, "step": 8856 }, { "epoch": 1.1267014374761481, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.67160415649414, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8820021152496338, "num_tokens": 337793802.0, "step": 8857 }, { "epoch": 1.1268286477547387, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.717838287353516, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8798658847808838, "num_tokens": 337828550.0, "step": 8858 }, { "epoch": 1.126955858033329, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.69316291809082, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8763594031333923, "num_tokens": 337862168.0, "step": 8859 }, { "epoch": 1.1270830683119195, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.740535736083984, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8628939390182495, "num_tokens": 337898837.0, "step": 8860 }, { "epoch": 1.12721027859051, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.66220474243164, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.853940486907959, "num_tokens": 337939608.0, "step": 8861 }, { "epoch": 1.1273374888691006, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.70315933227539, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.866753876209259, "num_tokens": 337983412.0, "step": 8862 }, { "epoch": 1.127464699147691, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.737051963806152, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.876401424407959, "num_tokens": 338022171.0, "step": 8863 }, { "epoch": 1.1275919094262816, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.682415008544922, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.864625096321106, "num_tokens": 338058838.0, "step": 8864 }, { "epoch": 1.1277191197048722, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.736474990844727, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.86375892162323, "num_tokens": 338095080.0, "step": 8865 }, { "epoch": 1.1278463299834627, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.724088668823242, "learning_rate": 1e-06, "loss": 0.4717, "mean_token_accuracy": 0.848764181137085, "num_tokens": 338134206.0, "step": 8866 }, { "epoch": 1.1279735402620532, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.663887023925781, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.873090922832489, "num_tokens": 338173794.0, "step": 8867 }, { "epoch": 1.1281007505406437, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.754311561584473, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8642862439155579, "num_tokens": 338206591.0, "step": 8868 }, { "epoch": 1.1282279608192343, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.649552345275879, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8789178729057312, "num_tokens": 338238728.0, "step": 8869 }, { "epoch": 1.1283551710978248, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.692802429199219, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8766535520553589, "num_tokens": 338282372.0, "step": 8870 }, { "epoch": 1.1284823813764153, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.694948196411133, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8625873923301697, "num_tokens": 338319175.0, "step": 8871 }, { "epoch": 1.1286095916550056, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.67845630645752, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8645871877670288, "num_tokens": 338357967.0, "step": 8872 }, { "epoch": 1.1287368019335962, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.712282180786133, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.874647319316864, "num_tokens": 338391431.0, "step": 8873 }, { "epoch": 1.1288640122121867, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.64312744140625, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8755841255187988, "num_tokens": 338426298.0, "step": 8874 }, { "epoch": 1.1289912224907772, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.72144889831543, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8494268655776978, "num_tokens": 338468147.0, "step": 8875 }, { "epoch": 1.1291184327693677, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.72021198272705, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8507933020591736, "num_tokens": 338517095.0, "step": 8876 }, { "epoch": 1.1292456430479583, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.71180534362793, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8745056390762329, "num_tokens": 338549925.0, "step": 8877 }, { "epoch": 1.1293728533265488, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.733497619628906, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8696093559265137, "num_tokens": 338583966.0, "step": 8878 }, { "epoch": 1.1295000636051393, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.620878219604492, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8715686798095703, "num_tokens": 338622715.0, "step": 8879 }, { "epoch": 1.1296272738837299, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.734305381774902, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8554217219352722, "num_tokens": 338659920.0, "step": 8880 }, { "epoch": 1.1297544841623204, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.656201362609863, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8617655038833618, "num_tokens": 338698183.0, "step": 8881 }, { "epoch": 1.129881694440911, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.680044174194336, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8556419610977173, "num_tokens": 338735853.0, "step": 8882 }, { "epoch": 1.1300089047195012, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.647233963012695, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8568481802940369, "num_tokens": 338771714.0, "step": 8883 }, { "epoch": 1.1301361149980917, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.687298774719238, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.88383549451828, "num_tokens": 338813823.0, "step": 8884 }, { "epoch": 1.1302633252766823, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.6459321975708, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8702977299690247, "num_tokens": 338858137.0, "step": 8885 }, { "epoch": 1.1303905355552728, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.659765243530273, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8564592003822327, "num_tokens": 338893653.0, "step": 8886 }, { "epoch": 1.1305177458338633, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.709978103637695, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.8483457565307617, "num_tokens": 338931230.0, "step": 8887 }, { "epoch": 1.1306449561124539, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.679287910461426, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.866101861000061, "num_tokens": 338966921.0, "step": 8888 }, { "epoch": 1.1307721663910444, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.788286209106445, "learning_rate": 1e-06, "loss": 0.4879, "mean_token_accuracy": 0.8480497598648071, "num_tokens": 339000957.0, "step": 8889 }, { "epoch": 1.130899376669635, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.681182861328125, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8597375750541687, "num_tokens": 339035876.0, "step": 8890 }, { "epoch": 1.1310265869482254, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.690390586853027, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8681354522705078, "num_tokens": 339073700.0, "step": 8891 }, { "epoch": 1.131153797226816, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.721429824829102, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8713914752006531, "num_tokens": 339108222.0, "step": 8892 }, { "epoch": 1.1312810075054065, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.644552230834961, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8779206275939941, "num_tokens": 339145749.0, "step": 8893 }, { "epoch": 1.131408217783997, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.729270935058594, "learning_rate": 1e-06, "loss": 0.4711, "mean_token_accuracy": 0.8535608649253845, "num_tokens": 339183852.0, "step": 8894 }, { "epoch": 1.1315354280625876, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.660005569458008, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8816555738449097, "num_tokens": 339218766.0, "step": 8895 }, { "epoch": 1.131662638341178, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.718350410461426, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8679147958755493, "num_tokens": 339259932.0, "step": 8896 }, { "epoch": 1.1317898486197684, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.705947875976562, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8704647421836853, "num_tokens": 339302201.0, "step": 8897 }, { "epoch": 1.131917058898359, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.677982330322266, "learning_rate": 1e-06, "loss": 0.5222, "mean_token_accuracy": 0.8382531404495239, "num_tokens": 339342091.0, "step": 8898 }, { "epoch": 1.1320442691769494, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.689919471740723, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8616924285888672, "num_tokens": 339379951.0, "step": 8899 }, { "epoch": 1.13217147945554, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.684582710266113, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8613765239715576, "num_tokens": 339422243.0, "step": 8900 }, { "epoch": 1.1322986897341305, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.656218528747559, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8813080787658691, "num_tokens": 339464969.0, "step": 8901 }, { "epoch": 1.132425900012721, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.668464660644531, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8710585832595825, "num_tokens": 339502382.0, "step": 8902 }, { "epoch": 1.1325531102913116, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.694873809814453, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.851778507232666, "num_tokens": 339542875.0, "step": 8903 }, { "epoch": 1.132680320569902, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.678019523620605, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8534915447235107, "num_tokens": 339580776.0, "step": 8904 }, { "epoch": 1.1328075308484926, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.646952629089355, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8793048858642578, "num_tokens": 339620195.0, "step": 8905 }, { "epoch": 1.1329347411270831, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.7027006149292, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8747777938842773, "num_tokens": 339659234.0, "step": 8906 }, { "epoch": 1.1330619514056737, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.659638404846191, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8570868372917175, "num_tokens": 339703259.0, "step": 8907 }, { "epoch": 1.133189161684264, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.623600959777832, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8737411499023438, "num_tokens": 339745901.0, "step": 8908 }, { "epoch": 1.1333163719628545, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.646480560302734, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8576548099517822, "num_tokens": 339790417.0, "step": 8909 }, { "epoch": 1.133443582241445, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.718267440795898, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8591563701629639, "num_tokens": 339827224.0, "step": 8910 }, { "epoch": 1.1335707925200356, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.675612449645996, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.861014187335968, "num_tokens": 339859932.0, "step": 8911 }, { "epoch": 1.133698002798626, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.73019790649414, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.854844331741333, "num_tokens": 339902416.0, "step": 8912 }, { "epoch": 1.1338252130772166, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.654720306396484, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8751258850097656, "num_tokens": 339940495.0, "step": 8913 }, { "epoch": 1.1339524233558071, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.688462257385254, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8494420051574707, "num_tokens": 339983010.0, "step": 8914 }, { "epoch": 1.1340796336343977, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.709787368774414, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8671730756759644, "num_tokens": 340020683.0, "step": 8915 }, { "epoch": 1.1342068439129882, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.675350189208984, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8579279184341431, "num_tokens": 340063179.0, "step": 8916 }, { "epoch": 1.1343340541915787, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.67091178894043, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8758350610733032, "num_tokens": 340100166.0, "step": 8917 }, { "epoch": 1.1344612644701693, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.682018280029297, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.862974226474762, "num_tokens": 340144325.0, "step": 8918 }, { "epoch": 1.1345884747487598, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.728510856628418, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8663893938064575, "num_tokens": 340180563.0, "step": 8919 }, { "epoch": 1.1347156850273503, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.671279907226562, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8528554439544678, "num_tokens": 340224621.0, "step": 8920 }, { "epoch": 1.1348428953059406, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.725129127502441, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8715229034423828, "num_tokens": 340264810.0, "step": 8921 }, { "epoch": 1.1349701055845312, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.659989356994629, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8796122670173645, "num_tokens": 340301158.0, "step": 8922 }, { "epoch": 1.1350973158631217, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.705500602722168, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8662926554679871, "num_tokens": 340339806.0, "step": 8923 }, { "epoch": 1.1352245261417122, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.653736114501953, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.872611403465271, "num_tokens": 340386412.0, "step": 8924 }, { "epoch": 1.1353517364203027, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.706351280212402, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8758328557014465, "num_tokens": 340431391.0, "step": 8925 }, { "epoch": 1.1354789466988933, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.749822616577148, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8701844811439514, "num_tokens": 340465904.0, "step": 8926 }, { "epoch": 1.1356061569774838, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.722987174987793, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8700836300849915, "num_tokens": 340504290.0, "step": 8927 }, { "epoch": 1.1357333672560743, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.748244285583496, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8607866168022156, "num_tokens": 340539100.0, "step": 8928 }, { "epoch": 1.1358605775346649, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.744978904724121, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8808383941650391, "num_tokens": 340569005.0, "step": 8929 }, { "epoch": 1.1359877878132554, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.678964614868164, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8660882711410522, "num_tokens": 340613476.0, "step": 8930 }, { "epoch": 1.136114998091846, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.697637557983398, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8596959114074707, "num_tokens": 340647982.0, "step": 8931 }, { "epoch": 1.1362422083704362, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.773492813110352, "learning_rate": 1e-06, "loss": 0.4906, "mean_token_accuracy": 0.8447756171226501, "num_tokens": 340683270.0, "step": 8932 }, { "epoch": 1.1363694186490267, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.701448440551758, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8705569505691528, "num_tokens": 340719955.0, "step": 8933 }, { "epoch": 1.1364966289276173, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.7118501663208, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.867863118648529, "num_tokens": 340753934.0, "step": 8934 }, { "epoch": 1.1366238392062078, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.693418502807617, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.872671365737915, "num_tokens": 340787545.0, "step": 8935 }, { "epoch": 1.1367510494847983, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.667010307312012, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8762502670288086, "num_tokens": 340826362.0, "step": 8936 }, { "epoch": 1.1368782597633889, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.685483932495117, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8709239363670349, "num_tokens": 340857871.0, "step": 8937 }, { "epoch": 1.1370054700419794, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.715925216674805, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8723960518836975, "num_tokens": 340897736.0, "step": 8938 }, { "epoch": 1.13713268032057, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.709272384643555, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8618441820144653, "num_tokens": 340935774.0, "step": 8939 }, { "epoch": 1.1372598905991604, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.729272842407227, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8703112602233887, "num_tokens": 340971497.0, "step": 8940 }, { "epoch": 1.137387100877751, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.68144702911377, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8674492239952087, "num_tokens": 341016209.0, "step": 8941 }, { "epoch": 1.1375143111563415, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.74584674835205, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8622143864631653, "num_tokens": 341052137.0, "step": 8942 }, { "epoch": 1.137641521434932, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.717500686645508, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8689913153648376, "num_tokens": 341083417.0, "step": 8943 }, { "epoch": 1.1377687317135226, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.728931427001953, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8699677586555481, "num_tokens": 341127798.0, "step": 8944 }, { "epoch": 1.137895941992113, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.654695510864258, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8529457449913025, "num_tokens": 341173485.0, "step": 8945 }, { "epoch": 1.1380231522707034, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.71938419342041, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8555203676223755, "num_tokens": 341212834.0, "step": 8946 }, { "epoch": 1.138150362549294, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.745899200439453, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8600406646728516, "num_tokens": 341246497.0, "step": 8947 }, { "epoch": 1.1382775728278844, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.699913024902344, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8516978025436401, "num_tokens": 341291498.0, "step": 8948 }, { "epoch": 1.138404783106475, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.732044219970703, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8604477643966675, "num_tokens": 341328104.0, "step": 8949 }, { "epoch": 1.1385319933850655, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.70946216583252, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8775699138641357, "num_tokens": 341359042.0, "step": 8950 }, { "epoch": 1.138659203663656, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.685842514038086, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8669924139976501, "num_tokens": 341403529.0, "step": 8951 }, { "epoch": 1.1387864139422466, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.649239540100098, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8731749653816223, "num_tokens": 341442509.0, "step": 8952 }, { "epoch": 1.138913624220837, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.710124015808105, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8633810877799988, "num_tokens": 341483015.0, "step": 8953 }, { "epoch": 1.1390408344994276, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.712250709533691, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.866658627986908, "num_tokens": 341519676.0, "step": 8954 }, { "epoch": 1.1391680447780181, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.669602394104004, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8666050434112549, "num_tokens": 341563483.0, "step": 8955 }, { "epoch": 1.1392952550566087, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.696578979492188, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8594223856925964, "num_tokens": 341598387.0, "step": 8956 }, { "epoch": 1.139422465335199, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.734413146972656, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8695639371871948, "num_tokens": 341640253.0, "step": 8957 }, { "epoch": 1.1395496756137895, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.684033393859863, "learning_rate": 1e-06, "loss": 0.3354, "mean_token_accuracy": 0.8932336568832397, "num_tokens": 341675788.0, "step": 8958 }, { "epoch": 1.13967688589238, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.761606216430664, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8665310144424438, "num_tokens": 341716917.0, "step": 8959 }, { "epoch": 1.1398040961709706, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.713003158569336, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8659132122993469, "num_tokens": 341758936.0, "step": 8960 }, { "epoch": 1.139931306449561, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.741196632385254, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8440343141555786, "num_tokens": 341797612.0, "step": 8961 }, { "epoch": 1.1400585167281516, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.69425106048584, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8623843193054199, "num_tokens": 341836061.0, "step": 8962 }, { "epoch": 1.1401857270067421, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.763583183288574, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8657186031341553, "num_tokens": 341875465.0, "step": 8963 }, { "epoch": 1.1403129372853327, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.702485084533691, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8608059883117676, "num_tokens": 341919795.0, "step": 8964 }, { "epoch": 1.1404401475639232, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.74619197845459, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8631306290626526, "num_tokens": 341953291.0, "step": 8965 }, { "epoch": 1.1405673578425137, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.756206512451172, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8851026892662048, "num_tokens": 341988364.0, "step": 8966 }, { "epoch": 1.1406945681211043, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.642033576965332, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8703631162643433, "num_tokens": 342028478.0, "step": 8967 }, { "epoch": 1.1408217783996948, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.769554138183594, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8738821744918823, "num_tokens": 342067348.0, "step": 8968 }, { "epoch": 1.1409489886782853, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.735662460327148, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8629879355430603, "num_tokens": 342105845.0, "step": 8969 }, { "epoch": 1.1410761989568756, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.717124938964844, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.867673397064209, "num_tokens": 342142072.0, "step": 8970 }, { "epoch": 1.1412034092354661, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.720721244812012, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.86910480260849, "num_tokens": 342176592.0, "step": 8971 }, { "epoch": 1.1413306195140567, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.734657287597656, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8694964647293091, "num_tokens": 342211896.0, "step": 8972 }, { "epoch": 1.1414578297926472, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.685043334960938, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8700146675109863, "num_tokens": 342256078.0, "step": 8973 }, { "epoch": 1.1415850400712377, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.699747085571289, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8690019845962524, "num_tokens": 342296766.0, "step": 8974 }, { "epoch": 1.1417122503498283, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.683287620544434, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.865851104259491, "num_tokens": 342333689.0, "step": 8975 }, { "epoch": 1.1418394606284188, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.739055633544922, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8776438236236572, "num_tokens": 342369605.0, "step": 8976 }, { "epoch": 1.1419666709070093, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.67422103881836, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8761435747146606, "num_tokens": 342410806.0, "step": 8977 }, { "epoch": 1.1420938811855998, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.735814094543457, "learning_rate": 1e-06, "loss": 0.3351, "mean_token_accuracy": 0.8918980360031128, "num_tokens": 342446957.0, "step": 8978 }, { "epoch": 1.1422210914641904, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.723709106445312, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8557215929031372, "num_tokens": 342480717.0, "step": 8979 }, { "epoch": 1.142348301742781, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.682090759277344, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8744823336601257, "num_tokens": 342520431.0, "step": 8980 }, { "epoch": 1.1424755120213712, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.719889640808105, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8636389970779419, "num_tokens": 342561735.0, "step": 8981 }, { "epoch": 1.1426027222999617, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.714509010314941, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8779726028442383, "num_tokens": 342593899.0, "step": 8982 }, { "epoch": 1.1427299325785523, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.743234634399414, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8660844564437866, "num_tokens": 342628044.0, "step": 8983 }, { "epoch": 1.1428571428571428, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.761066436767578, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8647924661636353, "num_tokens": 342665018.0, "step": 8984 }, { "epoch": 1.1429843531357333, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.744964599609375, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8556084632873535, "num_tokens": 342704832.0, "step": 8985 }, { "epoch": 1.1431115634143239, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.7568941116333, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.874024510383606, "num_tokens": 342741598.0, "step": 8986 }, { "epoch": 1.1432387736929144, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.709258079528809, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8620490431785583, "num_tokens": 342777302.0, "step": 8987 }, { "epoch": 1.143365983971505, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.762171745300293, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.846940279006958, "num_tokens": 342809578.0, "step": 8988 }, { "epoch": 1.1434931942500954, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.689842224121094, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8675270676612854, "num_tokens": 342845809.0, "step": 8989 }, { "epoch": 1.143620404528686, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.726489067077637, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8535701036453247, "num_tokens": 342886118.0, "step": 8990 }, { "epoch": 1.1437476148072765, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.725727081298828, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8626211285591125, "num_tokens": 342922517.0, "step": 8991 }, { "epoch": 1.143874825085867, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.777453422546387, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8606739044189453, "num_tokens": 342959741.0, "step": 8992 }, { "epoch": 1.1440020353644575, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.727231979370117, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8797358274459839, "num_tokens": 342991418.0, "step": 8993 }, { "epoch": 1.144129245643048, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.749825477600098, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8844464421272278, "num_tokens": 343024263.0, "step": 8994 }, { "epoch": 1.1442564559216384, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.728874206542969, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8675751686096191, "num_tokens": 343066388.0, "step": 8995 }, { "epoch": 1.144383666200229, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.711828231811523, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8680421710014343, "num_tokens": 343107631.0, "step": 8996 }, { "epoch": 1.1445108764788194, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.79319953918457, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.872808575630188, "num_tokens": 343143645.0, "step": 8997 }, { "epoch": 1.14463808675741, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.739287376403809, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8529279828071594, "num_tokens": 343182076.0, "step": 8998 }, { "epoch": 1.1447652970360005, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.699313163757324, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8848859071731567, "num_tokens": 343222808.0, "step": 8999 }, { "epoch": 1.144892507314591, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.71614933013916, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.864030122756958, "num_tokens": 343268229.0, "step": 9000 }, { "epoch": 1.1450197175931816, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.727261543273926, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8589988946914673, "num_tokens": 343304459.0, "step": 9001 }, { "epoch": 1.145146927871772, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.807045936584473, "learning_rate": 1e-06, "loss": 0.5401, "mean_token_accuracy": 0.8295201063156128, "num_tokens": 343343211.0, "step": 9002 }, { "epoch": 1.1452741381503626, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.704695701599121, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8622789978981018, "num_tokens": 343388498.0, "step": 9003 }, { "epoch": 1.1454013484289531, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.681558609008789, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8644959330558777, "num_tokens": 343430101.0, "step": 9004 }, { "epoch": 1.1455285587075437, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.74498176574707, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8546574115753174, "num_tokens": 343466454.0, "step": 9005 }, { "epoch": 1.145655768986134, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.733970642089844, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8817172646522522, "num_tokens": 343506960.0, "step": 9006 }, { "epoch": 1.1457829792647245, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.723672866821289, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8741171360015869, "num_tokens": 343541133.0, "step": 9007 }, { "epoch": 1.145910189543315, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.768819808959961, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8747124671936035, "num_tokens": 343580266.0, "step": 9008 }, { "epoch": 1.1460373998219056, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.705571174621582, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8641992211341858, "num_tokens": 343611691.0, "step": 9009 }, { "epoch": 1.146164610100496, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.754437446594238, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8725441694259644, "num_tokens": 343649355.0, "step": 9010 }, { "epoch": 1.1462918203790866, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.75036907196045, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8813654184341431, "num_tokens": 343681606.0, "step": 9011 }, { "epoch": 1.1464190306576771, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.72989559173584, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8774013519287109, "num_tokens": 343721592.0, "step": 9012 }, { "epoch": 1.1465462409362677, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.705937385559082, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.86420738697052, "num_tokens": 343764084.0, "step": 9013 }, { "epoch": 1.1466734512148582, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.702646255493164, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8678604960441589, "num_tokens": 343801213.0, "step": 9014 }, { "epoch": 1.1468006614934487, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.767969131469727, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8687089681625366, "num_tokens": 343839761.0, "step": 9015 }, { "epoch": 1.1469278717720393, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.720748901367188, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.866725504398346, "num_tokens": 343877792.0, "step": 9016 }, { "epoch": 1.1470550820506298, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.752208709716797, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8728336691856384, "num_tokens": 343915107.0, "step": 9017 }, { "epoch": 1.1471822923292203, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.768563270568848, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8718730807304382, "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": 11.776379585266113, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8496214151382446, "num_tokens": 343991058.0, "step": 9019 }, { "epoch": 1.1474367128864011, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.817838668823242, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8677492141723633, "num_tokens": 344030424.0, "step": 9020 }, { "epoch": 1.1475639231649917, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.766274452209473, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8692896366119385, "num_tokens": 344071385.0, "step": 9021 }, { "epoch": 1.1476911334435822, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.757061958312988, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8763654828071594, "num_tokens": 344109725.0, "step": 9022 }, { "epoch": 1.1478183437221727, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.693229675292969, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8750287294387817, "num_tokens": 344143947.0, "step": 9023 }, { "epoch": 1.1479455540007633, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.812418937683105, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.856751561164856, "num_tokens": 344179969.0, "step": 9024 }, { "epoch": 1.1480727642793538, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.73208999633789, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8785582780838013, "num_tokens": 344212279.0, "step": 9025 }, { "epoch": 1.1481999745579443, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.775736808776855, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8731321096420288, "num_tokens": 344248402.0, "step": 9026 }, { "epoch": 1.1483271848365348, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.720149040222168, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8642568588256836, "num_tokens": 344286557.0, "step": 9027 }, { "epoch": 1.1484543951151254, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.762557983398438, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.868087649345398, "num_tokens": 344331014.0, "step": 9028 }, { "epoch": 1.148581605393716, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.754878997802734, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8642602562904358, "num_tokens": 344373239.0, "step": 9029 }, { "epoch": 1.1487088156723062, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.764266967773438, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8818562030792236, "num_tokens": 344409871.0, "step": 9030 }, { "epoch": 1.1488360259508967, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.727693557739258, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8573887348175049, "num_tokens": 344452108.0, "step": 9031 }, { "epoch": 1.1489632362294873, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.801016807556152, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8615561723709106, "num_tokens": 344489798.0, "step": 9032 }, { "epoch": 1.1490904465080778, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.76692008972168, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8749158978462219, "num_tokens": 344528194.0, "step": 9033 }, { "epoch": 1.1492176567866683, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.862576484680176, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8528532385826111, "num_tokens": 344559699.0, "step": 9034 }, { "epoch": 1.1493448670652588, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.745455741882324, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8600367307662964, "num_tokens": 344598234.0, "step": 9035 }, { "epoch": 1.1494720773438494, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.787769317626953, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8611701726913452, "num_tokens": 344634780.0, "step": 9036 }, { "epoch": 1.14959928762244, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.742247581481934, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8713968992233276, "num_tokens": 344677167.0, "step": 9037 }, { "epoch": 1.1497264979010304, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.788991928100586, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8540478944778442, "num_tokens": 344713824.0, "step": 9038 }, { "epoch": 1.149853708179621, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.783592224121094, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.854714572429657, "num_tokens": 344755814.0, "step": 9039 }, { "epoch": 1.1499809184582115, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.729894638061523, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8816174268722534, "num_tokens": 344795022.0, "step": 9040 }, { "epoch": 1.150108128736802, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.742270469665527, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8765904903411865, "num_tokens": 344833906.0, "step": 9041 }, { "epoch": 1.1502353390153925, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.7471923828125, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8717567920684814, "num_tokens": 344875975.0, "step": 9042 }, { "epoch": 1.150362549293983, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.72814655303955, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8704460859298706, "num_tokens": 344915433.0, "step": 9043 }, { "epoch": 1.1504897595725734, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.76735782623291, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8708456158638, "num_tokens": 344949038.0, "step": 9044 }, { "epoch": 1.150616969851164, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.76150131225586, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8679928779602051, "num_tokens": 344988260.0, "step": 9045 }, { "epoch": 1.1507441801297544, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.759498596191406, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8796639442443848, "num_tokens": 345024355.0, "step": 9046 }, { "epoch": 1.150871390408345, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.737281799316406, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8764129877090454, "num_tokens": 345063041.0, "step": 9047 }, { "epoch": 1.1509986006869355, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.732023239135742, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8865225315093994, "num_tokens": 345097813.0, "step": 9048 }, { "epoch": 1.151125810965526, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.728597640991211, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8640962243080139, "num_tokens": 345134534.0, "step": 9049 }, { "epoch": 1.1512530212441165, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.811455726623535, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8710654377937317, "num_tokens": 345171018.0, "step": 9050 }, { "epoch": 1.151380231522707, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.72793197631836, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8787145614624023, "num_tokens": 345208006.0, "step": 9051 }, { "epoch": 1.1515074418012976, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.766094207763672, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8854971528053284, "num_tokens": 345246616.0, "step": 9052 }, { "epoch": 1.1516346520798881, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.750995635986328, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8838061094284058, "num_tokens": 345288095.0, "step": 9053 }, { "epoch": 1.1517618623584787, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.724366188049316, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8538976907730103, "num_tokens": 345333291.0, "step": 9054 }, { "epoch": 1.151889072637069, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.793914794921875, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8662549257278442, "num_tokens": 345375093.0, "step": 9055 }, { "epoch": 1.1520162829156595, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.776789665222168, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8609369397163391, "num_tokens": 345415547.0, "step": 9056 }, { "epoch": 1.15214349319425, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.726486206054688, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8716545104980469, "num_tokens": 345457340.0, "step": 9057 }, { "epoch": 1.1522707034728406, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.813859939575195, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8633419275283813, "num_tokens": 345494711.0, "step": 9058 }, { "epoch": 1.152397913751431, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.719715118408203, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8602473139762878, "num_tokens": 345534249.0, "step": 9059 }, { "epoch": 1.1525251240300216, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.797146797180176, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8666603565216064, "num_tokens": 345570832.0, "step": 9060 }, { "epoch": 1.1526523343086121, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.750306129455566, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8620463013648987, "num_tokens": 345609007.0, "step": 9061 }, { "epoch": 1.1527795445872027, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.741389274597168, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8785197138786316, "num_tokens": 345647979.0, "step": 9062 }, { "epoch": 1.1529067548657932, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.713353157043457, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8604040145874023, "num_tokens": 345681821.0, "step": 9063 }, { "epoch": 1.1530339651443837, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.79342269897461, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8577669262886047, "num_tokens": 345724274.0, "step": 9064 }, { "epoch": 1.1531611754229742, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.684895515441895, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8832365870475769, "num_tokens": 345763558.0, "step": 9065 }, { "epoch": 1.1532883857015648, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.784740447998047, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8620215654373169, "num_tokens": 345798486.0, "step": 9066 }, { "epoch": 1.1534155959801553, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.698905944824219, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8640343546867371, "num_tokens": 345839379.0, "step": 9067 }, { "epoch": 1.1535428062587456, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.766932487487793, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8638643026351929, "num_tokens": 345876675.0, "step": 9068 }, { "epoch": 1.1536700165373361, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.736663818359375, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8859909772872925, "num_tokens": 345913482.0, "step": 9069 }, { "epoch": 1.1537972268159267, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.7509183883667, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8633178472518921, "num_tokens": 345947671.0, "step": 9070 }, { "epoch": 1.1539244370945172, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.789907455444336, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8705084323883057, "num_tokens": 345981520.0, "step": 9071 }, { "epoch": 1.1540516473731077, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.782646179199219, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8500392436981201, "num_tokens": 346020050.0, "step": 9072 }, { "epoch": 1.1541788576516983, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.783061027526855, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8595343828201294, "num_tokens": 346050000.0, "step": 9073 }, { "epoch": 1.1543060679302888, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.831595420837402, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8611905574798584, "num_tokens": 346085038.0, "step": 9074 }, { "epoch": 1.1544332782088793, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.717537879943848, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8772700428962708, "num_tokens": 346125871.0, "step": 9075 }, { "epoch": 1.1545604884874698, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.794755935668945, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8758714199066162, "num_tokens": 346160460.0, "step": 9076 }, { "epoch": 1.1546876987660604, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.74846363067627, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.857662558555603, "num_tokens": 346201621.0, "step": 9077 }, { "epoch": 1.154814909044651, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.793776512145996, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8688676357269287, "num_tokens": 346242680.0, "step": 9078 }, { "epoch": 1.1549421193232412, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.78184700012207, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8578839302062988, "num_tokens": 346275774.0, "step": 9079 }, { "epoch": 1.1550693296018317, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.719987869262695, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.869587779045105, "num_tokens": 346317648.0, "step": 9080 }, { "epoch": 1.1551965398804223, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.792269706726074, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8700297474861145, "num_tokens": 346357736.0, "step": 9081 }, { "epoch": 1.1553237501590128, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.711528778076172, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8662564754486084, "num_tokens": 346397961.0, "step": 9082 }, { "epoch": 1.1554509604376033, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.795112609863281, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.878682017326355, "num_tokens": 346438268.0, "step": 9083 }, { "epoch": 1.1555781707161938, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.715603828430176, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8585388660430908, "num_tokens": 346479024.0, "step": 9084 }, { "epoch": 1.1557053809947844, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.771978378295898, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8750966787338257, "num_tokens": 346523526.0, "step": 9085 }, { "epoch": 1.155832591273375, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.711745262145996, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8757277727127075, "num_tokens": 346561985.0, "step": 9086 }, { "epoch": 1.1559598015519654, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.773818969726562, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.853967547416687, "num_tokens": 346605600.0, "step": 9087 }, { "epoch": 1.156087011830556, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.702402114868164, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8801020383834839, "num_tokens": 346640798.0, "step": 9088 }, { "epoch": 1.1562142221091465, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.805265426635742, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8617352247238159, "num_tokens": 346678483.0, "step": 9089 }, { "epoch": 1.156341432387737, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.710932731628418, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8839571475982666, "num_tokens": 346714604.0, "step": 9090 }, { "epoch": 1.1564686426663275, "ewc_loss": 0.025634765625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 11.776400566101074, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8788082599639893, "num_tokens": 346747203.0, "step": 9091 }, { "epoch": 1.156595852944918, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.68929672241211, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.866455078125, "num_tokens": 346790650.0, "step": 9092 }, { "epoch": 1.1567230632235084, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.757902145385742, "learning_rate": 1e-06, "loss": 0.3428, "mean_token_accuracy": 0.8907696008682251, "num_tokens": 346822872.0, "step": 9093 }, { "epoch": 1.156850273502099, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.77223014831543, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8740204572677612, "num_tokens": 346857999.0, "step": 9094 }, { "epoch": 1.1569774837806894, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.713858604431152, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8752642273902893, "num_tokens": 346891804.0, "step": 9095 }, { "epoch": 1.15710469405928, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.779300689697266, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8669690489768982, "num_tokens": 346931554.0, "step": 9096 }, { "epoch": 1.1572319043378705, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.78714656829834, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8547988533973694, "num_tokens": 346965214.0, "step": 9097 }, { "epoch": 1.157359114616461, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.770540237426758, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8794998526573181, "num_tokens": 347001219.0, "step": 9098 }, { "epoch": 1.1574863248950515, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.736292839050293, "learning_rate": 1e-06, "loss": 0.4649, "mean_token_accuracy": 0.8508464097976685, "num_tokens": 347043212.0, "step": 9099 }, { "epoch": 1.157613535173642, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.742481231689453, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8796365261077881, "num_tokens": 347082249.0, "step": 9100 }, { "epoch": 1.1577407454522326, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.755849838256836, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8657557964324951, "num_tokens": 347119286.0, "step": 9101 }, { "epoch": 1.1578679557308231, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.739198684692383, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.880753755569458, "num_tokens": 347155549.0, "step": 9102 }, { "epoch": 1.1579951660094137, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.741214752197266, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8698466420173645, "num_tokens": 347196846.0, "step": 9103 }, { "epoch": 1.158122376288004, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.803529739379883, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.882611870765686, "num_tokens": 347234028.0, "step": 9104 }, { "epoch": 1.1582495865665945, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.7179594039917, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8791124820709229, "num_tokens": 347276931.0, "step": 9105 }, { "epoch": 1.158376796845185, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.752829551696777, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8651170134544373, "num_tokens": 347317139.0, "step": 9106 }, { "epoch": 1.1585040071237755, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.748879432678223, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8710660338401794, "num_tokens": 347355718.0, "step": 9107 }, { "epoch": 1.158631217402366, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.757200241088867, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8584808707237244, "num_tokens": 347393403.0, "step": 9108 }, { "epoch": 1.1587584276809566, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.783862113952637, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8749692440032959, "num_tokens": 347433736.0, "step": 9109 }, { "epoch": 1.1588856379595471, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.72547435760498, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8691667318344116, "num_tokens": 347474574.0, "step": 9110 }, { "epoch": 1.1590128482381377, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.78730297088623, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8738700747489929, "num_tokens": 347518785.0, "step": 9111 }, { "epoch": 1.1591400585167282, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.728368759155273, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8557922840118408, "num_tokens": 347557518.0, "step": 9112 }, { "epoch": 1.1592672687953187, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.786704063415527, "learning_rate": 1e-06, "loss": 0.4977, "mean_token_accuracy": 0.8386670351028442, "num_tokens": 347600094.0, "step": 9113 }, { "epoch": 1.1593944790739092, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.7905855178833, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.867081880569458, "num_tokens": 347633709.0, "step": 9114 }, { "epoch": 1.1595216893524998, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.747143745422363, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8753308057785034, "num_tokens": 347670027.0, "step": 9115 }, { "epoch": 1.1596488996310903, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.76663875579834, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8732261061668396, "num_tokens": 347707488.0, "step": 9116 }, { "epoch": 1.1597761099096806, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.763710975646973, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8627719879150391, "num_tokens": 347746030.0, "step": 9117 }, { "epoch": 1.1599033201882711, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.783891677856445, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8613520264625549, "num_tokens": 347783409.0, "step": 9118 }, { "epoch": 1.1600305304668617, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.753215789794922, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8752883672714233, "num_tokens": 347821335.0, "step": 9119 }, { "epoch": 1.1601577407454522, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.7665433883667, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8793581128120422, "num_tokens": 347857590.0, "step": 9120 }, { "epoch": 1.1602849510240427, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.735940933227539, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8788704872131348, "num_tokens": 347897000.0, "step": 9121 }, { "epoch": 1.1604121613026332, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.808849334716797, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8683947324752808, "num_tokens": 347928947.0, "step": 9122 }, { "epoch": 1.1605393715812238, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.830683708190918, "learning_rate": 1e-06, "loss": 0.5224, "mean_token_accuracy": 0.8378498554229736, "num_tokens": 347959294.0, "step": 9123 }, { "epoch": 1.1606665818598143, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.796327590942383, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8734922409057617, "num_tokens": 347993103.0, "step": 9124 }, { "epoch": 1.1607937921384048, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.733258247375488, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8802805542945862, "num_tokens": 348028412.0, "step": 9125 }, { "epoch": 1.1609210024169954, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.782976150512695, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8506608605384827, "num_tokens": 348071429.0, "step": 9126 }, { "epoch": 1.161048212695586, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.790268898010254, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8729497194290161, "num_tokens": 348113222.0, "step": 9127 }, { "epoch": 1.1611754229741762, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.77061653137207, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8664158582687378, "num_tokens": 348150188.0, "step": 9128 }, { "epoch": 1.1613026332527667, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.768280982971191, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8714469075202942, "num_tokens": 348188547.0, "step": 9129 }, { "epoch": 1.1614298435313573, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.780802726745605, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8632533550262451, "num_tokens": 348228624.0, "step": 9130 }, { "epoch": 1.1615570538099478, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.756915092468262, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8644539713859558, "num_tokens": 348264932.0, "step": 9131 }, { "epoch": 1.1616842640885383, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.712230682373047, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.870064377784729, "num_tokens": 348307680.0, "step": 9132 }, { "epoch": 1.1618114743671288, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.79074478149414, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8827884197235107, "num_tokens": 348346449.0, "step": 9133 }, { "epoch": 1.1619386846457194, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.681730270385742, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.879059374332428, "num_tokens": 348386393.0, "step": 9134 }, { "epoch": 1.16206589492431, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.819101333618164, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8614410161972046, "num_tokens": 348429458.0, "step": 9135 }, { "epoch": 1.1621931052029004, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.754557609558105, "learning_rate": 1e-06, "loss": 0.4534, "mean_token_accuracy": 0.8560580611228943, "num_tokens": 348471365.0, "step": 9136 }, { "epoch": 1.162320315481491, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.752671241760254, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8540890216827393, "num_tokens": 348513425.0, "step": 9137 }, { "epoch": 1.1624475257600815, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.762248992919922, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8665692210197449, "num_tokens": 348551181.0, "step": 9138 }, { "epoch": 1.162574736038672, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.769777297973633, "learning_rate": 1e-06, "loss": 0.3351, "mean_token_accuracy": 0.8887450695037842, "num_tokens": 348583611.0, "step": 9139 }, { "epoch": 1.1627019463172625, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.781063079833984, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8658341765403748, "num_tokens": 348624281.0, "step": 9140 }, { "epoch": 1.162829156595853, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.734129905700684, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8665376901626587, "num_tokens": 348661416.0, "step": 9141 }, { "epoch": 1.1629563668744434, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.806761741638184, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8472707867622375, "num_tokens": 348703907.0, "step": 9142 }, { "epoch": 1.163083577153034, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.718305587768555, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8804259896278381, "num_tokens": 348735617.0, "step": 9143 }, { "epoch": 1.1632107874316244, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.80870246887207, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8651484251022339, "num_tokens": 348773552.0, "step": 9144 }, { "epoch": 1.163337997710215, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.76646900177002, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8690198063850403, "num_tokens": 348811274.0, "step": 9145 }, { "epoch": 1.1634652079888055, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.819432258605957, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8561191558837891, "num_tokens": 348847773.0, "step": 9146 }, { "epoch": 1.163592418267396, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.809273719787598, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8595848679542542, "num_tokens": 348886356.0, "step": 9147 }, { "epoch": 1.1637196285459865, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.77353572845459, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8785965442657471, "num_tokens": 348922892.0, "step": 9148 }, { "epoch": 1.163846838824577, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.79775619506836, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8871273398399353, "num_tokens": 348954924.0, "step": 9149 }, { "epoch": 1.1639740491031676, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.80002212524414, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.874296247959137, "num_tokens": 348997335.0, "step": 9150 }, { "epoch": 1.1641012593817581, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.761878967285156, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8571678400039673, "num_tokens": 349038221.0, "step": 9151 }, { "epoch": 1.1642284696603487, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.802008628845215, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8642828464508057, "num_tokens": 349076559.0, "step": 9152 }, { "epoch": 1.164355679938939, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.8231782913208, "learning_rate": 1e-06, "loss": 0.4788, "mean_token_accuracy": 0.8467697501182556, "num_tokens": 349112264.0, "step": 9153 }, { "epoch": 1.1644828902175295, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.801531791687012, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8802130222320557, "num_tokens": 349146571.0, "step": 9154 }, { "epoch": 1.16461010049612, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.781375885009766, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8757801055908203, "num_tokens": 349183674.0, "step": 9155 }, { "epoch": 1.1647373107747105, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.772887229919434, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8724892139434814, "num_tokens": 349227042.0, "step": 9156 }, { "epoch": 1.164864521053301, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.817461013793945, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8634276390075684, "num_tokens": 349264072.0, "step": 9157 }, { "epoch": 1.1649917313318916, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.78410816192627, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8687980771064758, "num_tokens": 349304427.0, "step": 9158 }, { "epoch": 1.1651189416104821, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.756892204284668, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.867862343788147, "num_tokens": 349346708.0, "step": 9159 }, { "epoch": 1.1652461518890727, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.807113647460938, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8668856620788574, "num_tokens": 349387716.0, "step": 9160 }, { "epoch": 1.1653733621676632, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.775856971740723, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8652595281600952, "num_tokens": 349425757.0, "step": 9161 }, { "epoch": 1.1655005724462537, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.797776222229004, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8753188848495483, "num_tokens": 349472097.0, "step": 9162 }, { "epoch": 1.1656277827248442, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.828445434570312, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8658255338668823, "num_tokens": 349513809.0, "step": 9163 }, { "epoch": 1.1657549930034348, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.815528869628906, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8837375640869141, "num_tokens": 349544906.0, "step": 9164 }, { "epoch": 1.1658822032820253, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.792884826660156, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8633534908294678, "num_tokens": 349582529.0, "step": 9165 }, { "epoch": 1.1660094135606156, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.780529022216797, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8628156185150146, "num_tokens": 349624871.0, "step": 9166 }, { "epoch": 1.1661366238392061, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.816736221313477, "learning_rate": 1e-06, "loss": 0.5117, "mean_token_accuracy": 0.8325989842414856, "num_tokens": 349660783.0, "step": 9167 }, { "epoch": 1.1662638341177967, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.804909706115723, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8783589601516724, "num_tokens": 349698430.0, "step": 9168 }, { "epoch": 1.1663910443963872, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.805399894714355, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8603520393371582, "num_tokens": 349734981.0, "step": 9169 }, { "epoch": 1.1665182546749777, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.827402114868164, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8627879023551941, "num_tokens": 349774267.0, "step": 9170 }, { "epoch": 1.1666454649535682, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.799335479736328, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8643264174461365, "num_tokens": 349811400.0, "step": 9171 }, { "epoch": 1.1667726752321588, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.832746505737305, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8597360849380493, "num_tokens": 349850031.0, "step": 9172 }, { "epoch": 1.1668998855107493, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.821515083312988, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8708950281143188, "num_tokens": 349889967.0, "step": 9173 }, { "epoch": 1.1670270957893398, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.831411361694336, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8616514205932617, "num_tokens": 349930276.0, "step": 9174 }, { "epoch": 1.1671543060679304, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.844505310058594, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8634455800056458, "num_tokens": 349969889.0, "step": 9175 }, { "epoch": 1.1672815163465209, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.859899520874023, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8654637336730957, "num_tokens": 350005729.0, "step": 9176 }, { "epoch": 1.1674087266251112, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.841437339782715, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8751112222671509, "num_tokens": 350043130.0, "step": 9177 }, { "epoch": 1.1675359369037017, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.77507209777832, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8856531381607056, "num_tokens": 350081233.0, "step": 9178 }, { "epoch": 1.1676631471822922, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.848370552062988, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8542150855064392, "num_tokens": 350119403.0, "step": 9179 }, { "epoch": 1.1677903574608828, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.762443542480469, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8562315702438354, "num_tokens": 350159099.0, "step": 9180 }, { "epoch": 1.1679175677394733, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.837187767028809, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8679236173629761, "num_tokens": 350195642.0, "step": 9181 }, { "epoch": 1.1680447780180638, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.787014961242676, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8741497993469238, "num_tokens": 350236721.0, "step": 9182 }, { "epoch": 1.1681719882966544, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.7841215133667, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8642162084579468, "num_tokens": 350272691.0, "step": 9183 }, { "epoch": 1.168299198575245, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.817127227783203, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8666510581970215, "num_tokens": 350318299.0, "step": 9184 }, { "epoch": 1.1684264088538354, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.746416091918945, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8672150373458862, "num_tokens": 350360973.0, "step": 9185 }, { "epoch": 1.168553619132426, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.799788475036621, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8759810328483582, "num_tokens": 350400516.0, "step": 9186 }, { "epoch": 1.1686808294110165, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.744972229003906, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8636335134506226, "num_tokens": 350438147.0, "step": 9187 }, { "epoch": 1.168808039689607, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.839329719543457, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8724532127380371, "num_tokens": 350481604.0, "step": 9188 }, { "epoch": 1.1689352499681975, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.778794288635254, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8884165287017822, "num_tokens": 350518225.0, "step": 9189 }, { "epoch": 1.169062460246788, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.851278305053711, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8667261600494385, "num_tokens": 350553752.0, "step": 9190 }, { "epoch": 1.1691896705253784, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.764923095703125, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8726719617843628, "num_tokens": 350593676.0, "step": 9191 }, { "epoch": 1.169316880803969, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.797933578491211, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8670899868011475, "num_tokens": 350633618.0, "step": 9192 }, { "epoch": 1.1694440910825594, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.800601959228516, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.865513801574707, "num_tokens": 350678221.0, "step": 9193 }, { "epoch": 1.16957130136115, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.741146087646484, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8780179023742676, "num_tokens": 350718258.0, "step": 9194 }, { "epoch": 1.1696985116397405, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.769770622253418, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.878319501876831, "num_tokens": 350757370.0, "step": 9195 }, { "epoch": 1.169825721918331, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.848723411560059, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8683756589889526, "num_tokens": 350801394.0, "step": 9196 }, { "epoch": 1.1699529321969215, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.787409782409668, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8599436283111572, "num_tokens": 350840635.0, "step": 9197 }, { "epoch": 1.170080142475512, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.764555931091309, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8843458890914917, "num_tokens": 350879616.0, "step": 9198 }, { "epoch": 1.1702073527541026, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.772900581359863, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8678205609321594, "num_tokens": 350921710.0, "step": 9199 }, { "epoch": 1.1703345630326931, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.772371292114258, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8559220433235168, "num_tokens": 350964218.0, "step": 9200 }, { "epoch": 1.1704617733112836, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.765881538391113, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8774566054344177, "num_tokens": 351004511.0, "step": 9201 }, { "epoch": 1.170588983589874, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.780649185180664, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8636082410812378, "num_tokens": 351044583.0, "step": 9202 }, { "epoch": 1.1707161938684645, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.761527061462402, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8504472374916077, "num_tokens": 351083613.0, "step": 9203 }, { "epoch": 1.170843404147055, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.795605659484863, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8601826429367065, "num_tokens": 351117412.0, "step": 9204 }, { "epoch": 1.1709706144256455, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.779964447021484, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8541020750999451, "num_tokens": 351154779.0, "step": 9205 }, { "epoch": 1.171097824704236, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.84656047821045, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8784633874893188, "num_tokens": 351191412.0, "step": 9206 }, { "epoch": 1.1712250349828266, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.748927116394043, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8709015846252441, "num_tokens": 351235465.0, "step": 9207 }, { "epoch": 1.1713522452614171, "ewc_loss": 0.02587890625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 11.77729606628418, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8733813166618347, "num_tokens": 351279326.0, "step": 9208 }, { "epoch": 1.1714794555400077, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.740391731262207, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8850513696670532, "num_tokens": 351316678.0, "step": 9209 }, { "epoch": 1.1716066658185982, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.800640106201172, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8647254705429077, "num_tokens": 351352641.0, "step": 9210 }, { "epoch": 1.1717338760971887, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.849440574645996, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8587468862533569, "num_tokens": 351385845.0, "step": 9211 }, { "epoch": 1.1718610863757792, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.789446830749512, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8741751909255981, "num_tokens": 351425220.0, "step": 9212 }, { "epoch": 1.1719882966543698, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.794500350952148, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8625695109367371, "num_tokens": 351467515.0, "step": 9213 }, { "epoch": 1.1721155069329603, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.82365608215332, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.865243136882782, "num_tokens": 351507571.0, "step": 9214 }, { "epoch": 1.1722427172115506, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.757588386535645, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8656840324401855, "num_tokens": 351546150.0, "step": 9215 }, { "epoch": 1.1723699274901411, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.748743057250977, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8614363670349121, "num_tokens": 351588574.0, "step": 9216 }, { "epoch": 1.1724971377687317, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.771690368652344, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8688850402832031, "num_tokens": 351626914.0, "step": 9217 }, { "epoch": 1.1726243480473222, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.830288887023926, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8578579425811768, "num_tokens": 351663574.0, "step": 9218 }, { "epoch": 1.1727515583259127, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.804962158203125, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.851357102394104, "num_tokens": 351703275.0, "step": 9219 }, { "epoch": 1.1728787686045032, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.749801635742188, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8583630919456482, "num_tokens": 351745713.0, "step": 9220 }, { "epoch": 1.1730059788830938, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.814359664916992, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8532372713088989, "num_tokens": 351780789.0, "step": 9221 }, { "epoch": 1.1731331891616843, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.80117416381836, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8777459859848022, "num_tokens": 351820352.0, "step": 9222 }, { "epoch": 1.1732603994402748, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.853803634643555, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8748407363891602, "num_tokens": 351860011.0, "step": 9223 }, { "epoch": 1.1733876097188654, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.766569137573242, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8620421886444092, "num_tokens": 351898742.0, "step": 9224 }, { "epoch": 1.1735148199974559, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.828798294067383, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.854804277420044, "num_tokens": 351940201.0, "step": 9225 }, { "epoch": 1.1736420302760462, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.822173118591309, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8660111427307129, "num_tokens": 351971041.0, "step": 9226 }, { "epoch": 1.1737692405546367, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.819470405578613, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8507609367370605, "num_tokens": 352013002.0, "step": 9227 }, { "epoch": 1.1738964508332272, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.808838844299316, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8661585450172424, "num_tokens": 352051435.0, "step": 9228 }, { "epoch": 1.1740236611118178, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.810093879699707, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8733919262886047, "num_tokens": 352093295.0, "step": 9229 }, { "epoch": 1.1741508713904083, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.778334617614746, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8635439872741699, "num_tokens": 352136401.0, "step": 9230 }, { "epoch": 1.1742780816689988, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.81391429901123, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8785903453826904, "num_tokens": 352178696.0, "step": 9231 }, { "epoch": 1.1744052919475894, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.843379974365234, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8486055135726929, "num_tokens": 352214759.0, "step": 9232 }, { "epoch": 1.1745325022261799, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.796759605407715, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8804017901420593, "num_tokens": 352250520.0, "step": 9233 }, { "epoch": 1.1746597125047704, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.853745460510254, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8659743070602417, "num_tokens": 352288773.0, "step": 9234 }, { "epoch": 1.174786922783361, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.77529525756836, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8747718930244446, "num_tokens": 352326548.0, "step": 9235 }, { "epoch": 1.1749141330619515, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.846268653869629, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8765522241592407, "num_tokens": 352366537.0, "step": 9236 }, { "epoch": 1.175041343340542, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.8250732421875, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.850649356842041, "num_tokens": 352402390.0, "step": 9237 }, { "epoch": 1.1751685536191325, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.809622764587402, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8648892641067505, "num_tokens": 352442357.0, "step": 9238 }, { "epoch": 1.175295763897723, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.846364974975586, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8651615381240845, "num_tokens": 352476884.0, "step": 9239 }, { "epoch": 1.1754229741763134, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.780308723449707, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8674744963645935, "num_tokens": 352513448.0, "step": 9240 }, { "epoch": 1.175550184454904, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.792463302612305, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8862618803977966, "num_tokens": 352548420.0, "step": 9241 }, { "epoch": 1.1756773947334944, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.804824829101562, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8711569309234619, "num_tokens": 352592157.0, "step": 9242 }, { "epoch": 1.175804605012085, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.823955535888672, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8589236736297607, "num_tokens": 352626427.0, "step": 9243 }, { "epoch": 1.1759318152906755, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.819450378417969, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8768372535705566, "num_tokens": 352667075.0, "step": 9244 }, { "epoch": 1.176059025569266, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.773237228393555, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8652573227882385, "num_tokens": 352707114.0, "step": 9245 }, { "epoch": 1.1761862358478565, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.868492126464844, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8728501796722412, "num_tokens": 352738064.0, "step": 9246 }, { "epoch": 1.176313446126447, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.781256675720215, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.85219407081604, "num_tokens": 352779964.0, "step": 9247 }, { "epoch": 1.1764406564050376, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.851086616516113, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8752370476722717, "num_tokens": 352819574.0, "step": 9248 }, { "epoch": 1.1765678666836281, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.82886791229248, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8756776452064514, "num_tokens": 352859151.0, "step": 9249 }, { "epoch": 1.1766950769622184, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.823751449584961, "learning_rate": 1e-06, "loss": 0.5062, "mean_token_accuracy": 0.8459550142288208, "num_tokens": 352901119.0, "step": 9250 }, { "epoch": 1.176822287240809, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.851422309875488, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8665794134140015, "num_tokens": 352945878.0, "step": 9251 }, { "epoch": 1.1769494975193995, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.839065551757812, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.861616849899292, "num_tokens": 352979819.0, "step": 9252 }, { "epoch": 1.17707670779799, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.790070533752441, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8564724922180176, "num_tokens": 353023552.0, "step": 9253 }, { "epoch": 1.1772039180765805, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.817608833312988, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8685215711593628, "num_tokens": 353058178.0, "step": 9254 }, { "epoch": 1.177331128355171, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.782554626464844, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8768401145935059, "num_tokens": 353091428.0, "step": 9255 }, { "epoch": 1.1774583386337616, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.787662506103516, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8704229593276978, "num_tokens": 353131943.0, "step": 9256 }, { "epoch": 1.1775855489123521, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.790206909179688, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8605670928955078, "num_tokens": 353171213.0, "step": 9257 }, { "epoch": 1.1777127591909426, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.88630199432373, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.875211775302887, "num_tokens": 353205057.0, "step": 9258 }, { "epoch": 1.1778399694695332, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.79317855834961, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8532872796058655, "num_tokens": 353251164.0, "step": 9259 }, { "epoch": 1.1779671797481237, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.865116119384766, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.872141420841217, "num_tokens": 353288202.0, "step": 9260 }, { "epoch": 1.1780943900267142, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.850813865661621, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.867117166519165, "num_tokens": 353327396.0, "step": 9261 }, { "epoch": 1.1782216003053048, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.798622131347656, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8562759160995483, "num_tokens": 353366512.0, "step": 9262 }, { "epoch": 1.1783488105838953, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.824383735656738, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8589644432067871, "num_tokens": 353408546.0, "step": 9263 }, { "epoch": 1.1784760208624856, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.840873718261719, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.874845027923584, "num_tokens": 353444950.0, "step": 9264 }, { "epoch": 1.1786032311410761, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.827296257019043, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8740590810775757, "num_tokens": 353483306.0, "step": 9265 }, { "epoch": 1.1787304414196667, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.825484275817871, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8662149310112, "num_tokens": 353523038.0, "step": 9266 }, { "epoch": 1.1788576516982572, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.833894729614258, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8764625787734985, "num_tokens": 353564266.0, "step": 9267 }, { "epoch": 1.1789848619768477, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.799639701843262, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8710250854492188, "num_tokens": 353603558.0, "step": 9268 }, { "epoch": 1.1791120722554382, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.812935829162598, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8744367957115173, "num_tokens": 353648097.0, "step": 9269 }, { "epoch": 1.1792392825340288, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.868134498596191, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8521281480789185, "num_tokens": 353688970.0, "step": 9270 }, { "epoch": 1.1793664928126193, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.807092666625977, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8568342328071594, "num_tokens": 353732505.0, "step": 9271 }, { "epoch": 1.1794937030912098, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.860177993774414, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8756533265113831, "num_tokens": 353767460.0, "step": 9272 }, { "epoch": 1.1796209133698004, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.875802040100098, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.876398503780365, "num_tokens": 353801951.0, "step": 9273 }, { "epoch": 1.1797481236483909, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.856719970703125, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8744708895683289, "num_tokens": 353836815.0, "step": 9274 }, { "epoch": 1.1798753339269812, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.817353248596191, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.859483003616333, "num_tokens": 353876296.0, "step": 9275 }, { "epoch": 1.1800025442055717, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 11.838727951049805, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.850818395614624, "num_tokens": 353911792.0, "step": 9276 }, { "epoch": 1.1801297544841622, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.811630249023438, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8561644554138184, "num_tokens": 353949532.0, "step": 9277 }, { "epoch": 1.1802569647627528, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.818793296813965, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8547383546829224, "num_tokens": 353988622.0, "step": 9278 }, { "epoch": 1.1803841750413433, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.786666870117188, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.870343804359436, "num_tokens": 354028136.0, "step": 9279 }, { "epoch": 1.1805113853199338, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.81375789642334, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.869658887386322, "num_tokens": 354065029.0, "step": 9280 }, { "epoch": 1.1806385955985244, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.76877498626709, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8780820369720459, "num_tokens": 354099279.0, "step": 9281 }, { "epoch": 1.1807658058771149, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.806148529052734, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8620574474334717, "num_tokens": 354138870.0, "step": 9282 }, { "epoch": 1.1808930161557054, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.783628463745117, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8546257615089417, "num_tokens": 354176498.0, "step": 9283 }, { "epoch": 1.181020226434296, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.847023010253906, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8478047251701355, "num_tokens": 354210063.0, "step": 9284 }, { "epoch": 1.1811474367128865, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.8531494140625, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8591715693473816, "num_tokens": 354248095.0, "step": 9285 }, { "epoch": 1.181274646991477, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.786283493041992, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8632887601852417, "num_tokens": 354286897.0, "step": 9286 }, { "epoch": 1.1814018572700675, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.811261177062988, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8756262063980103, "num_tokens": 354330685.0, "step": 9287 }, { "epoch": 1.181529067548658, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.824974060058594, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8584328293800354, "num_tokens": 354371120.0, "step": 9288 }, { "epoch": 1.1816562778272484, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.845191955566406, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8650264739990234, "num_tokens": 354410825.0, "step": 9289 }, { "epoch": 1.1817834881058389, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.79806137084961, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8638287782669067, "num_tokens": 354453623.0, "step": 9290 }, { "epoch": 1.1819106983844294, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.832030296325684, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8672192096710205, "num_tokens": 354491854.0, "step": 9291 }, { "epoch": 1.18203790866302, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.843693733215332, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8526718616485596, "num_tokens": 354530788.0, "step": 9292 }, { "epoch": 1.1821651189416105, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.785318374633789, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8861463665962219, "num_tokens": 354568576.0, "step": 9293 }, { "epoch": 1.182292329220201, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.805619239807129, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8670096397399902, "num_tokens": 354611307.0, "step": 9294 }, { "epoch": 1.1824195394987915, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.84572982788086, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8736233711242676, "num_tokens": 354648810.0, "step": 9295 }, { "epoch": 1.182546749777382, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.850654602050781, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8686764240264893, "num_tokens": 354688411.0, "step": 9296 }, { "epoch": 1.1826739600559726, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.815244674682617, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8655315637588501, "num_tokens": 354734700.0, "step": 9297 }, { "epoch": 1.1828011703345631, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.862987518310547, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8632968664169312, "num_tokens": 354765517.0, "step": 9298 }, { "epoch": 1.1829283806131534, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.830556869506836, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8710946440696716, "num_tokens": 354808563.0, "step": 9299 }, { "epoch": 1.183055590891744, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.849752426147461, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8490478992462158, "num_tokens": 354846618.0, "step": 9300 }, { "epoch": 1.1831828011703345, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.881799697875977, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8690621852874756, "num_tokens": 354881293.0, "step": 9301 }, { "epoch": 1.183310011448925, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.775774955749512, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8651458621025085, "num_tokens": 354925239.0, "step": 9302 }, { "epoch": 1.1834372217275155, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.811528205871582, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.853866457939148, "num_tokens": 354967836.0, "step": 9303 }, { "epoch": 1.183564432006106, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.866936683654785, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8515911102294922, "num_tokens": 355003386.0, "step": 9304 }, { "epoch": 1.1836916422846966, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.829852104187012, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.8457998037338257, "num_tokens": 355043391.0, "step": 9305 }, { "epoch": 1.1838188525632871, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.770819664001465, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8682739734649658, "num_tokens": 355082820.0, "step": 9306 }, { "epoch": 1.1839460628418776, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.847124099731445, "learning_rate": 1e-06, "loss": 0.3503, "mean_token_accuracy": 0.8874905109405518, "num_tokens": 355117531.0, "step": 9307 }, { "epoch": 1.1840732731204682, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.857111930847168, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8551563024520874, "num_tokens": 355159485.0, "step": 9308 }, { "epoch": 1.1842004833990587, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.792556762695312, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8623383045196533, "num_tokens": 355206051.0, "step": 9309 }, { "epoch": 1.1843276936776492, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.864486694335938, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8585399985313416, "num_tokens": 355242933.0, "step": 9310 }, { "epoch": 1.1844549039562398, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.809274673461914, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8800311088562012, "num_tokens": 355286311.0, "step": 9311 }, { "epoch": 1.1845821142348303, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.844658851623535, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8659077286720276, "num_tokens": 355327844.0, "step": 9312 }, { "epoch": 1.1847093245134206, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.801788330078125, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8754897713661194, "num_tokens": 355368977.0, "step": 9313 }, { "epoch": 1.1848365347920111, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.845207214355469, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8568219542503357, "num_tokens": 355411479.0, "step": 9314 }, { "epoch": 1.1849637450706016, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.808183670043945, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.88374924659729, "num_tokens": 355447370.0, "step": 9315 }, { "epoch": 1.1850909553491922, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.772808074951172, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8640464544296265, "num_tokens": 355490335.0, "step": 9316 }, { "epoch": 1.1852181656277827, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.86466121673584, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8674123287200928, "num_tokens": 355527843.0, "step": 9317 }, { "epoch": 1.1853453759063732, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.817338943481445, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8637789487838745, "num_tokens": 355566094.0, "step": 9318 }, { "epoch": 1.1854725861849638, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.817729949951172, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8750556707382202, "num_tokens": 355609505.0, "step": 9319 }, { "epoch": 1.1855997964635543, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.8306884765625, "learning_rate": 1e-06, "loss": 0.4908, "mean_token_accuracy": 0.8426135182380676, "num_tokens": 355650930.0, "step": 9320 }, { "epoch": 1.1857270067421448, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.83633804321289, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8765611052513123, "num_tokens": 355684075.0, "step": 9321 }, { "epoch": 1.1858542170207353, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.801156044006348, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8715246915817261, "num_tokens": 355723567.0, "step": 9322 }, { "epoch": 1.1859814272993259, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.812490463256836, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8533055782318115, "num_tokens": 355764915.0, "step": 9323 }, { "epoch": 1.1861086375779162, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.856725692749023, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8685587048530579, "num_tokens": 355804950.0, "step": 9324 }, { "epoch": 1.1862358478565067, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.830301284790039, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8752360939979553, "num_tokens": 355844155.0, "step": 9325 }, { "epoch": 1.1863630581350972, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.80617904663086, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8829649686813354, "num_tokens": 355883971.0, "step": 9326 }, { "epoch": 1.1864902684136878, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.832446098327637, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.861667811870575, "num_tokens": 355922271.0, "step": 9327 }, { "epoch": 1.1866174786922783, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.834733009338379, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8664909601211548, "num_tokens": 355961087.0, "step": 9328 }, { "epoch": 1.1867446889708688, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.875046730041504, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8704225420951843, "num_tokens": 355999553.0, "step": 9329 }, { "epoch": 1.1868718992494594, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.780730247497559, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8723936676979065, "num_tokens": 356038503.0, "step": 9330 }, { "epoch": 1.1869991095280499, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.858657836914062, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8766283988952637, "num_tokens": 356078538.0, "step": 9331 }, { "epoch": 1.1871263198066404, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.832444190979004, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8790112733840942, "num_tokens": 356115031.0, "step": 9332 }, { "epoch": 1.187253530085231, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.84714412689209, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8808591365814209, "num_tokens": 356154751.0, "step": 9333 }, { "epoch": 1.1873807403638215, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.832720756530762, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.851733922958374, "num_tokens": 356192105.0, "step": 9334 }, { "epoch": 1.187507950642412, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.855624198913574, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8567017912864685, "num_tokens": 356224964.0, "step": 9335 }, { "epoch": 1.1876351609210025, "ewc_loss": 0.026123046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.89212417602539, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8714244961738586, "num_tokens": 356261565.0, "step": 9336 }, { "epoch": 1.187762371199593, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.829741477966309, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8684654235839844, "num_tokens": 356301886.0, "step": 9337 }, { "epoch": 1.1878895814781834, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.86146068572998, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8755011558532715, "num_tokens": 356339481.0, "step": 9338 }, { "epoch": 1.1880167917567739, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.804743766784668, "learning_rate": 1e-06, "loss": 0.4866, "mean_token_accuracy": 0.8438152074813843, "num_tokens": 356387038.0, "step": 9339 }, { "epoch": 1.1881440020353644, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.851751327514648, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8637866973876953, "num_tokens": 356433256.0, "step": 9340 }, { "epoch": 1.188271212313955, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.827728271484375, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8564026355743408, "num_tokens": 356471282.0, "step": 9341 }, { "epoch": 1.1883984225925455, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.861281394958496, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.873608410358429, "num_tokens": 356508951.0, "step": 9342 }, { "epoch": 1.188525632871136, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.907149314880371, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8691716194152832, "num_tokens": 356544487.0, "step": 9343 }, { "epoch": 1.1886528431497265, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.873459815979004, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.872957706451416, "num_tokens": 356580873.0, "step": 9344 }, { "epoch": 1.188780053428317, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.815174102783203, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8654001355171204, "num_tokens": 356613622.0, "step": 9345 }, { "epoch": 1.1889072637069076, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.882352828979492, "learning_rate": 1e-06, "loss": 0.4777, "mean_token_accuracy": 0.8444414138793945, "num_tokens": 356660058.0, "step": 9346 }, { "epoch": 1.189034473985498, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.872218132019043, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8657527565956116, "num_tokens": 356701614.0, "step": 9347 }, { "epoch": 1.1891616842640884, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.785365104675293, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8813877105712891, "num_tokens": 356737105.0, "step": 9348 }, { "epoch": 1.189288894542679, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.875727653503418, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8578956723213196, "num_tokens": 356779382.0, "step": 9349 }, { "epoch": 1.1894161048212695, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.853443145751953, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8619512319564819, "num_tokens": 356813836.0, "step": 9350 }, { "epoch": 1.18954331509986, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.866642951965332, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8577135801315308, "num_tokens": 356852591.0, "step": 9351 }, { "epoch": 1.1896705253784505, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.883732795715332, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8648014664649963, "num_tokens": 356889940.0, "step": 9352 }, { "epoch": 1.189797735657041, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.818167686462402, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8697474002838135, "num_tokens": 356927878.0, "step": 9353 }, { "epoch": 1.1899249459356316, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.925922393798828, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8762956857681274, "num_tokens": 356963339.0, "step": 9354 }, { "epoch": 1.1900521562142221, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.838696479797363, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8706972599029541, "num_tokens": 357002999.0, "step": 9355 }, { "epoch": 1.1901793664928126, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.850406646728516, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8695884943008423, "num_tokens": 357044305.0, "step": 9356 }, { "epoch": 1.1903065767714032, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.850785255432129, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.868564248085022, "num_tokens": 357082197.0, "step": 9357 }, { "epoch": 1.1904337870499937, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.88972282409668, "learning_rate": 1e-06, "loss": 0.4798, "mean_token_accuracy": 0.8482012748718262, "num_tokens": 357119541.0, "step": 9358 }, { "epoch": 1.1905609973285842, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.820361137390137, "learning_rate": 1e-06, "loss": 0.3485, "mean_token_accuracy": 0.8880671262741089, "num_tokens": 357159176.0, "step": 9359 }, { "epoch": 1.1906882076071748, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.889765739440918, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8663421273231506, "num_tokens": 357199204.0, "step": 9360 }, { "epoch": 1.1908154178857653, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.85453987121582, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8807304501533508, "num_tokens": 357236900.0, "step": 9361 }, { "epoch": 1.1909426281643556, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.866634368896484, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.867097795009613, "num_tokens": 357271622.0, "step": 9362 }, { "epoch": 1.1910698384429461, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.808124542236328, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8771673440933228, "num_tokens": 357311145.0, "step": 9363 }, { "epoch": 1.1911970487215366, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.899465560913086, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8840852379798889, "num_tokens": 357344609.0, "step": 9364 }, { "epoch": 1.1913242590001272, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.880160331726074, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8644418716430664, "num_tokens": 357378896.0, "step": 9365 }, { "epoch": 1.1914514692787177, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.820878982543945, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8708366751670837, "num_tokens": 357424730.0, "step": 9366 }, { "epoch": 1.1915786795573082, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.869275093078613, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.8488715887069702, "num_tokens": 357461152.0, "step": 9367 }, { "epoch": 1.1917058898358988, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.849517822265625, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8603349924087524, "num_tokens": 357502305.0, "step": 9368 }, { "epoch": 1.1918331001144893, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.895312309265137, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8621139526367188, "num_tokens": 357539187.0, "step": 9369 }, { "epoch": 1.1919603103930798, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.881919860839844, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8565547466278076, "num_tokens": 357569520.0, "step": 9370 }, { "epoch": 1.1920875206716703, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.919313430786133, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8659292459487915, "num_tokens": 357605253.0, "step": 9371 }, { "epoch": 1.1922147309502609, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.832574844360352, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8776555061340332, "num_tokens": 357639701.0, "step": 9372 }, { "epoch": 1.1923419412288512, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.90124797821045, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8748072385787964, "num_tokens": 357675611.0, "step": 9373 }, { "epoch": 1.1924691515074417, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.86939525604248, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8785223960876465, "num_tokens": 357708521.0, "step": 9374 }, { "epoch": 1.1925963617860322, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.860281944274902, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8576617240905762, "num_tokens": 357749970.0, "step": 9375 }, { "epoch": 1.1927235720646228, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.874292373657227, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8769472241401672, "num_tokens": 357783287.0, "step": 9376 }, { "epoch": 1.1928507823432133, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.971814155578613, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8683357238769531, "num_tokens": 357818883.0, "step": 9377 }, { "epoch": 1.1929779926218038, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.791473388671875, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.880133867263794, "num_tokens": 357862862.0, "step": 9378 }, { "epoch": 1.1931052029003943, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.900279998779297, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8559079766273499, "num_tokens": 357901574.0, "step": 9379 }, { "epoch": 1.1932324131789849, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.858511924743652, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8660306930541992, "num_tokens": 357941822.0, "step": 9380 }, { "epoch": 1.1933596234575754, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.897798538208008, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8599557280540466, "num_tokens": 357982820.0, "step": 9381 }, { "epoch": 1.193486833736166, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.861237525939941, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8649477362632751, "num_tokens": 358020491.0, "step": 9382 }, { "epoch": 1.1936140440147565, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.847357749938965, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8688866496086121, "num_tokens": 358059775.0, "step": 9383 }, { "epoch": 1.193741254293347, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.833553314208984, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8746023178100586, "num_tokens": 358098672.0, "step": 9384 }, { "epoch": 1.1938684645719375, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.886584281921387, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8588041067123413, "num_tokens": 358140850.0, "step": 9385 }, { "epoch": 1.193995674850528, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.806045532226562, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8737608194351196, "num_tokens": 358175503.0, "step": 9386 }, { "epoch": 1.1941228851291183, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.87403392791748, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8703117370605469, "num_tokens": 358210615.0, "step": 9387 }, { "epoch": 1.1942500954077089, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.843305587768555, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8692793846130371, "num_tokens": 358246605.0, "step": 9388 }, { "epoch": 1.1943773056862994, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.857331275939941, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8629736304283142, "num_tokens": 358283136.0, "step": 9389 }, { "epoch": 1.19450451596489, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.820377349853516, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8749623894691467, "num_tokens": 358325437.0, "step": 9390 }, { "epoch": 1.1946317262434805, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.892915725708008, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.866640031337738, "num_tokens": 358365360.0, "step": 9391 }, { "epoch": 1.194758936522071, "ewc_loss": 0.0263671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.811087608337402, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8843808174133301, "num_tokens": 358406502.0, "step": 9392 }, { "epoch": 1.1948861468006615, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.841443061828613, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8621146082878113, "num_tokens": 358437220.0, "step": 9393 }, { "epoch": 1.195013357079252, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.886187553405762, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8558143377304077, "num_tokens": 358475314.0, "step": 9394 }, { "epoch": 1.1951405673578426, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.87829303741455, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8777979016304016, "num_tokens": 358520777.0, "step": 9395 }, { "epoch": 1.195267777636433, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.842141151428223, "learning_rate": 1e-06, "loss": 0.3556, "mean_token_accuracy": 0.886150062084198, "num_tokens": 358557973.0, "step": 9396 }, { "epoch": 1.1953949879150234, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.855956077575684, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8744138479232788, "num_tokens": 358596795.0, "step": 9397 }, { "epoch": 1.195522198193614, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.837630271911621, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8676382303237915, "num_tokens": 358632146.0, "step": 9398 }, { "epoch": 1.1956494084722045, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.825757026672363, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8749181032180786, "num_tokens": 358672194.0, "step": 9399 }, { "epoch": 1.195776618750795, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.90709400177002, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8748323917388916, "num_tokens": 358710192.0, "step": 9400 }, { "epoch": 1.1959038290293855, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.885210037231445, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8545337319374084, "num_tokens": 358749284.0, "step": 9401 }, { "epoch": 1.196031039307976, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.88752555847168, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8640651106834412, "num_tokens": 358782830.0, "step": 9402 }, { "epoch": 1.1961582495865666, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.902559280395508, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8653451800346375, "num_tokens": 358820696.0, "step": 9403 }, { "epoch": 1.196285459865157, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.88208293914795, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8746207356452942, "num_tokens": 358853616.0, "step": 9404 }, { "epoch": 1.1964126701437476, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.86316204071045, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.875869870185852, "num_tokens": 358893420.0, "step": 9405 }, { "epoch": 1.1965398804223382, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.84492301940918, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8693705797195435, "num_tokens": 358930704.0, "step": 9406 }, { "epoch": 1.1966670907009287, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.87911319732666, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8798413276672363, "num_tokens": 358967730.0, "step": 9407 }, { "epoch": 1.1967943009795192, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.900367736816406, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.879905104637146, "num_tokens": 359007241.0, "step": 9408 }, { "epoch": 1.1969215112581097, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.889849662780762, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8554900884628296, "num_tokens": 359045573.0, "step": 9409 }, { "epoch": 1.1970487215367003, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.890607833862305, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8702472448348999, "num_tokens": 359083420.0, "step": 9410 }, { "epoch": 1.1971759318152906, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.878169059753418, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8804319500923157, "num_tokens": 359117555.0, "step": 9411 }, { "epoch": 1.1973031420938811, "ewc_loss": 0.026123046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 11.866602897644043, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8638099431991577, "num_tokens": 359154597.0, "step": 9412 }, { "epoch": 1.1974303523724716, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.939668655395508, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8628799319267273, "num_tokens": 359190074.0, "step": 9413 }, { "epoch": 1.1975575626510622, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.85696029663086, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8558861613273621, "num_tokens": 359232711.0, "step": 9414 }, { "epoch": 1.1976847729296527, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.912674903869629, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8681372404098511, "num_tokens": 359265582.0, "step": 9415 }, { "epoch": 1.1978119832082432, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.870222091674805, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.875257134437561, "num_tokens": 359303177.0, "step": 9416 }, { "epoch": 1.1979391934868338, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.865604400634766, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8777530789375305, "num_tokens": 359338712.0, "step": 9417 }, { "epoch": 1.1980664037654243, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.910738945007324, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8762459754943848, "num_tokens": 359372849.0, "step": 9418 }, { "epoch": 1.1981936140440148, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.855982780456543, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8730252385139465, "num_tokens": 359407981.0, "step": 9419 }, { "epoch": 1.1983208243226053, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.861174583435059, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8468968868255615, "num_tokens": 359450180.0, "step": 9420 }, { "epoch": 1.1984480346011959, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.806057929992676, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8668707609176636, "num_tokens": 359487909.0, "step": 9421 }, { "epoch": 1.1985752448797862, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.855703353881836, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8802026510238647, "num_tokens": 359522609.0, "step": 9422 }, { "epoch": 1.1987024551583767, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.864150047302246, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8648736476898193, "num_tokens": 359560735.0, "step": 9423 }, { "epoch": 1.1988296654369672, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.83691692352295, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8616139888763428, "num_tokens": 359602994.0, "step": 9424 }, { "epoch": 1.1989568757155578, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.81523323059082, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8674867153167725, "num_tokens": 359638864.0, "step": 9425 }, { "epoch": 1.1990840859941483, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.877106666564941, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8761109113693237, "num_tokens": 359674202.0, "step": 9426 }, { "epoch": 1.1992112962727388, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.891658782958984, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8733387589454651, "num_tokens": 359710408.0, "step": 9427 }, { "epoch": 1.1993385065513293, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.859658241271973, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8751763105392456, "num_tokens": 359749068.0, "step": 9428 }, { "epoch": 1.1994657168299199, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.85590934753418, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.868576169013977, "num_tokens": 359788449.0, "step": 9429 }, { "epoch": 1.1995929271085104, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.861018180847168, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8628875017166138, "num_tokens": 359824167.0, "step": 9430 }, { "epoch": 1.199720137387101, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.921476364135742, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8636296391487122, "num_tokens": 359861476.0, "step": 9431 }, { "epoch": 1.1998473476656915, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.867143630981445, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8666937947273254, "num_tokens": 359906463.0, "step": 9432 }, { "epoch": 1.199974557944282, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.867286682128906, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8712577223777771, "num_tokens": 359940576.0, "step": 9433 }, { "epoch": 1.2001017682228725, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.94675064086914, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8781505823135376, "num_tokens": 359973616.0, "step": 9434 }, { "epoch": 1.200228978501463, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.838754653930664, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8801027536392212, "num_tokens": 360011059.0, "step": 9435 }, { "epoch": 1.2003561887800533, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.888978958129883, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8775374293327332, "num_tokens": 360048307.0, "step": 9436 }, { "epoch": 1.2004833990586439, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.84676456451416, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8679299354553223, "num_tokens": 360086946.0, "step": 9437 }, { "epoch": 1.2006106093372344, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.816429138183594, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8723281025886536, "num_tokens": 360121908.0, "step": 9438 }, { "epoch": 1.200737819615825, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.853414535522461, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8594425320625305, "num_tokens": 360164349.0, "step": 9439 }, { "epoch": 1.2008650298944155, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.902616500854492, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.8836196660995483, "num_tokens": 360197660.0, "step": 9440 }, { "epoch": 1.200992240173006, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.824926376342773, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8466501832008362, "num_tokens": 360240297.0, "step": 9441 }, { "epoch": 1.2011194504515965, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.876749992370605, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8581935167312622, "num_tokens": 360281566.0, "step": 9442 }, { "epoch": 1.201246660730187, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.896381378173828, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8737710118293762, "num_tokens": 360313927.0, "step": 9443 }, { "epoch": 1.2013738710087776, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.81177043914795, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8724616765975952, "num_tokens": 360349369.0, "step": 9444 }, { "epoch": 1.201501081287368, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.884587287902832, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8670071363449097, "num_tokens": 360386996.0, "step": 9445 }, { "epoch": 1.2016282915659584, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.88131046295166, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8721268177032471, "num_tokens": 360424883.0, "step": 9446 }, { "epoch": 1.201755501844549, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.902602195739746, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8740504384040833, "num_tokens": 360457992.0, "step": 9447 }, { "epoch": 1.2018827121231395, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.853068351745605, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8533014059066772, "num_tokens": 360497934.0, "step": 9448 }, { "epoch": 1.20200992240173, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.896259307861328, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8657460808753967, "num_tokens": 360533928.0, "step": 9449 }, { "epoch": 1.2021371326803205, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.902312278747559, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8706225156784058, "num_tokens": 360571076.0, "step": 9450 }, { "epoch": 1.202264342958911, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.823576927185059, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8600560426712036, "num_tokens": 360611957.0, "step": 9451 }, { "epoch": 1.2023915532375016, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.888324737548828, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8615560531616211, "num_tokens": 360647823.0, "step": 9452 }, { "epoch": 1.202518763516092, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.864683151245117, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8594346642494202, "num_tokens": 360688341.0, "step": 9453 }, { "epoch": 1.2026459737946826, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.893328666687012, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8596231937408447, "num_tokens": 360730417.0, "step": 9454 }, { "epoch": 1.2027731840732732, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.857308387756348, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8492224216461182, "num_tokens": 360771931.0, "step": 9455 }, { "epoch": 1.2029003943518637, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.854150772094727, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.865978479385376, "num_tokens": 360811460.0, "step": 9456 }, { "epoch": 1.2030276046304542, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.820119857788086, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8799272775650024, "num_tokens": 360849476.0, "step": 9457 }, { "epoch": 1.2031548149090447, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.896790504455566, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8571463823318481, "num_tokens": 360891178.0, "step": 9458 }, { "epoch": 1.2032820251876353, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.83505630493164, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8637363910675049, "num_tokens": 360930503.0, "step": 9459 }, { "epoch": 1.2034092354662256, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.863502502441406, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8649518489837646, "num_tokens": 360963730.0, "step": 9460 }, { "epoch": 1.203536445744816, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.826300621032715, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8541039228439331, "num_tokens": 361003796.0, "step": 9461 }, { "epoch": 1.2036636560234066, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.848155975341797, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8682290315628052, "num_tokens": 361044818.0, "step": 9462 }, { "epoch": 1.2037908663019972, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.836173057556152, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8779107332229614, "num_tokens": 361080692.0, "step": 9463 }, { "epoch": 1.2039180765805877, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.884157180786133, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8585473895072937, "num_tokens": 361116609.0, "step": 9464 }, { "epoch": 1.2040452868591782, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.881780624389648, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8601937294006348, "num_tokens": 361154559.0, "step": 9465 }, { "epoch": 1.2041724971377687, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.815751075744629, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8764358758926392, "num_tokens": 361191005.0, "step": 9466 }, { "epoch": 1.2042997074163593, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.864569664001465, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.865558385848999, "num_tokens": 361230191.0, "step": 9467 }, { "epoch": 1.2044269176949498, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.82604694366455, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8613272905349731, "num_tokens": 361269132.0, "step": 9468 }, { "epoch": 1.2045541279735403, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.901421546936035, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8604159355163574, "num_tokens": 361306674.0, "step": 9469 }, { "epoch": 1.2046813382521309, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.80791187286377, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8692969679832458, "num_tokens": 361348309.0, "step": 9470 }, { "epoch": 1.2048085485307212, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.891153335571289, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8627117872238159, "num_tokens": 361386824.0, "step": 9471 }, { "epoch": 1.2049357588093117, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.902974128723145, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8489053845405579, "num_tokens": 361426590.0, "step": 9472 }, { "epoch": 1.2050629690879022, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.906519889831543, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8536382913589478, "num_tokens": 361463259.0, "step": 9473 }, { "epoch": 1.2051901793664928, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.883170127868652, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8769755959510803, "num_tokens": 361498750.0, "step": 9474 }, { "epoch": 1.2053173896450833, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.86512279510498, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8568209409713745, "num_tokens": 361537101.0, "step": 9475 }, { "epoch": 1.2054445999236738, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.886652946472168, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.870395302772522, "num_tokens": 361572130.0, "step": 9476 }, { "epoch": 1.2055718102022643, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.82298469543457, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8836895227432251, "num_tokens": 361608484.0, "step": 9477 }, { "epoch": 1.2056990204808549, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.887048721313477, "learning_rate": 1e-06, "loss": 0.353, "mean_token_accuracy": 0.8844480514526367, "num_tokens": 361650361.0, "step": 9478 }, { "epoch": 1.2058262307594454, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.9030179977417, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8749951124191284, "num_tokens": 361680670.0, "step": 9479 }, { "epoch": 1.205953441038036, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.861312866210938, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8733797073364258, "num_tokens": 361723504.0, "step": 9480 }, { "epoch": 1.2060806513166265, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.852441787719727, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8738683462142944, "num_tokens": 361765850.0, "step": 9481 }, { "epoch": 1.206207861595217, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.857452392578125, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8584883213043213, "num_tokens": 361804375.0, "step": 9482 }, { "epoch": 1.2063350718738075, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.87804126739502, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8649067282676697, "num_tokens": 361834714.0, "step": 9483 }, { "epoch": 1.206462282152398, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.869263648986816, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8678075075149536, "num_tokens": 361877568.0, "step": 9484 }, { "epoch": 1.2065894924309883, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.869258880615234, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8634400963783264, "num_tokens": 361914785.0, "step": 9485 }, { "epoch": 1.2067167027095789, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.874256134033203, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8750761151313782, "num_tokens": 361956781.0, "step": 9486 }, { "epoch": 1.2068439129881694, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.836503028869629, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8774481415748596, "num_tokens": 361997408.0, "step": 9487 }, { "epoch": 1.20697112326676, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.878332138061523, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8678976893424988, "num_tokens": 362030226.0, "step": 9488 }, { "epoch": 1.2070983335453505, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.834342002868652, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8779840469360352, "num_tokens": 362068370.0, "step": 9489 }, { "epoch": 1.207225543823941, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.85887622833252, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8757835626602173, "num_tokens": 362108569.0, "step": 9490 }, { "epoch": 1.2073527541025315, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.856437683105469, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.860650360584259, "num_tokens": 362140476.0, "step": 9491 }, { "epoch": 1.207479964381122, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.914737701416016, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8762824535369873, "num_tokens": 362171132.0, "step": 9492 }, { "epoch": 1.2076071746597126, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.853704452514648, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8650914430618286, "num_tokens": 362204791.0, "step": 9493 }, { "epoch": 1.207734384938303, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.844030380249023, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8715589046478271, "num_tokens": 362242360.0, "step": 9494 }, { "epoch": 1.2078615952168934, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.89349365234375, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8704580664634705, "num_tokens": 362276558.0, "step": 9495 }, { "epoch": 1.207988805495484, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.90123176574707, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8538929224014282, "num_tokens": 362307055.0, "step": 9496 }, { "epoch": 1.2081160157740745, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.846271514892578, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8786555528640747, "num_tokens": 362342433.0, "step": 9497 }, { "epoch": 1.208243226052665, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.888179779052734, "learning_rate": 1e-06, "loss": 0.3361, "mean_token_accuracy": 0.8890305161476135, "num_tokens": 362377008.0, "step": 9498 }, { "epoch": 1.2083704363312555, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.90202808380127, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8821993470191956, "num_tokens": 362412089.0, "step": 9499 }, { "epoch": 1.208497646609846, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.853943824768066, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8599298596382141, "num_tokens": 362455817.0, "step": 9500 }, { "epoch": 1.2086248568884366, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.835721969604492, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8680762648582458, "num_tokens": 362498859.0, "step": 9501 }, { "epoch": 1.208752067167027, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.849833488464355, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8674435615539551, "num_tokens": 362542346.0, "step": 9502 }, { "epoch": 1.2088792774456176, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.85690975189209, "learning_rate": 1e-06, "loss": 0.338, "mean_token_accuracy": 0.8873356580734253, "num_tokens": 362575360.0, "step": 9503 }, { "epoch": 1.2090064877242082, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.862293243408203, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8691977262496948, "num_tokens": 362614737.0, "step": 9504 }, { "epoch": 1.2091336980027987, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.873815536499023, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8628392219543457, "num_tokens": 362653756.0, "step": 9505 }, { "epoch": 1.2092609082813892, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.83554744720459, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8699203133583069, "num_tokens": 362690745.0, "step": 9506 }, { "epoch": 1.2093881185599797, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.873451232910156, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.877240777015686, "num_tokens": 362726328.0, "step": 9507 }, { "epoch": 1.2095153288385703, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.865644454956055, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.871462345123291, "num_tokens": 362764285.0, "step": 9508 }, { "epoch": 1.2096425391171606, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.870002746582031, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8497499227523804, "num_tokens": 362794697.0, "step": 9509 }, { "epoch": 1.209769749395751, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.882375717163086, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8669605255126953, "num_tokens": 362831618.0, "step": 9510 }, { "epoch": 1.2098969596743416, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.837261199951172, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8757244348526001, "num_tokens": 362866404.0, "step": 9511 }, { "epoch": 1.2100241699529322, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.865656852722168, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8845677971839905, "num_tokens": 362902554.0, "step": 9512 }, { "epoch": 1.2101513802315227, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.903533935546875, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8630527853965759, "num_tokens": 362944169.0, "step": 9513 }, { "epoch": 1.2102785905101132, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.875783920288086, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8478178381919861, "num_tokens": 362983789.0, "step": 9514 }, { "epoch": 1.2104058007887037, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.906320571899414, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8744306564331055, "num_tokens": 363021551.0, "step": 9515 }, { "epoch": 1.2105330110672943, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.837149620056152, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8518607020378113, "num_tokens": 363064951.0, "step": 9516 }, { "epoch": 1.2106602213458848, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.832720756530762, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8678609132766724, "num_tokens": 363110876.0, "step": 9517 }, { "epoch": 1.2107874316244753, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.877167701721191, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8651872873306274, "num_tokens": 363155310.0, "step": 9518 }, { "epoch": 1.2109146419030659, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.86988639831543, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8793863654136658, "num_tokens": 363193166.0, "step": 9519 }, { "epoch": 1.2110418521816562, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.891748428344727, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8503763675689697, "num_tokens": 363225163.0, "step": 9520 }, { "epoch": 1.2111690624602467, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.91972827911377, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8674269318580627, "num_tokens": 363264720.0, "step": 9521 }, { "epoch": 1.2112962727388372, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.873315811157227, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8527510762214661, "num_tokens": 363305029.0, "step": 9522 }, { "epoch": 1.2114234830174277, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.846726417541504, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8744058012962341, "num_tokens": 363344128.0, "step": 9523 }, { "epoch": 1.2115506932960183, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.872458457946777, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8666629791259766, "num_tokens": 363385903.0, "step": 9524 }, { "epoch": 1.2116779035746088, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.910070419311523, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8693687915802002, "num_tokens": 363425216.0, "step": 9525 }, { "epoch": 1.2118051138531993, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.863870620727539, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8793767094612122, "num_tokens": 363462691.0, "step": 9526 }, { "epoch": 1.2119323241317899, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.912620544433594, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8542150259017944, "num_tokens": 363495849.0, "step": 9527 }, { "epoch": 1.2120595344103804, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.891962051391602, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8668359518051147, "num_tokens": 363535268.0, "step": 9528 }, { "epoch": 1.212186744688971, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.872200965881348, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8506344556808472, "num_tokens": 363573523.0, "step": 9529 }, { "epoch": 1.2123139549675614, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.880355834960938, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8673964738845825, "num_tokens": 363607545.0, "step": 9530 }, { "epoch": 1.212441165246152, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.860625267028809, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8702136874198914, "num_tokens": 363648456.0, "step": 9531 }, { "epoch": 1.2125683755247425, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.864997863769531, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8769396543502808, "num_tokens": 363688726.0, "step": 9532 }, { "epoch": 1.212695585803333, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.872139930725098, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8610595464706421, "num_tokens": 363728557.0, "step": 9533 }, { "epoch": 1.2128227960819233, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.84834098815918, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8676294684410095, "num_tokens": 363764191.0, "step": 9534 }, { "epoch": 1.2129500063605139, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.961654663085938, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8566473126411438, "num_tokens": 363798126.0, "step": 9535 }, { "epoch": 1.2130772166391044, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.898836135864258, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8513860702514648, "num_tokens": 363836037.0, "step": 9536 }, { "epoch": 1.213204426917695, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.933852195739746, "learning_rate": 1e-06, "loss": 0.3499, "mean_token_accuracy": 0.8882275819778442, "num_tokens": 363872359.0, "step": 9537 }, { "epoch": 1.2133316371962855, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.906182289123535, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.862619936466217, "num_tokens": 363914206.0, "step": 9538 }, { "epoch": 1.213458847474876, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.933484077453613, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.86906898021698, "num_tokens": 363956147.0, "step": 9539 }, { "epoch": 1.2135860577534665, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.923093795776367, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8850471377372742, "num_tokens": 363986328.0, "step": 9540 }, { "epoch": 1.213713268032057, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.961333274841309, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8653140068054199, "num_tokens": 364024905.0, "step": 9541 }, { "epoch": 1.2138404783106476, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.912580490112305, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.869122326374054, "num_tokens": 364068628.0, "step": 9542 }, { "epoch": 1.213967688589238, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.939783096313477, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8647302389144897, "num_tokens": 364104261.0, "step": 9543 }, { "epoch": 1.2140948988678284, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.90770435333252, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8647664785385132, "num_tokens": 364147524.0, "step": 9544 }, { "epoch": 1.214222109146419, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.933614730834961, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8695623874664307, "num_tokens": 364187715.0, "step": 9545 }, { "epoch": 1.2143493194250095, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.929604530334473, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8782042264938354, "num_tokens": 364226081.0, "step": 9546 }, { "epoch": 1.2144765297036, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.907005310058594, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8688989281654358, "num_tokens": 364268390.0, "step": 9547 }, { "epoch": 1.2146037399821905, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.914558410644531, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.870768666267395, "num_tokens": 364310176.0, "step": 9548 }, { "epoch": 1.214730950260781, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.884754180908203, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8798712491989136, "num_tokens": 364345652.0, "step": 9549 }, { "epoch": 1.2148581605393716, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.934724807739258, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8645675182342529, "num_tokens": 364383810.0, "step": 9550 }, { "epoch": 1.214985370817962, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.884632110595703, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8723297119140625, "num_tokens": 364421169.0, "step": 9551 }, { "epoch": 1.2151125810965526, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.930586814880371, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8618216514587402, "num_tokens": 364452552.0, "step": 9552 }, { "epoch": 1.2152397913751432, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.914549827575684, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.865966260433197, "num_tokens": 364486779.0, "step": 9553 }, { "epoch": 1.2153670016537337, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.88521671295166, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.878153383731842, "num_tokens": 364526237.0, "step": 9554 }, { "epoch": 1.2154942119323242, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.904359817504883, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8744669556617737, "num_tokens": 364568446.0, "step": 9555 }, { "epoch": 1.2156214222109147, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.888927459716797, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8755975365638733, "num_tokens": 364604448.0, "step": 9556 }, { "epoch": 1.2157486324895053, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.873985290527344, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8652538061141968, "num_tokens": 364642707.0, "step": 9557 }, { "epoch": 1.2158758427680956, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.980820655822754, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8640397787094116, "num_tokens": 364678428.0, "step": 9558 }, { "epoch": 1.216003053046686, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.880212783813477, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8812698125839233, "num_tokens": 364718915.0, "step": 9559 }, { "epoch": 1.2161302633252766, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.901969909667969, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8841369152069092, "num_tokens": 364756030.0, "step": 9560 }, { "epoch": 1.2162574736038672, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.914190292358398, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8651437163352966, "num_tokens": 364796158.0, "step": 9561 }, { "epoch": 1.2163846838824577, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.94519329071045, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8760545253753662, "num_tokens": 364832722.0, "step": 9562 }, { "epoch": 1.2165118941610482, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.860379219055176, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8659781217575073, "num_tokens": 364874120.0, "step": 9563 }, { "epoch": 1.2166391044396387, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.909180641174316, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8701562881469727, "num_tokens": 364913166.0, "step": 9564 }, { "epoch": 1.2167663147182293, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.91043472290039, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8493976593017578, "num_tokens": 364949586.0, "step": 9565 }, { "epoch": 1.2168935249968198, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.897390365600586, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8502942323684692, "num_tokens": 364989070.0, "step": 9566 }, { "epoch": 1.2170207352754103, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.90354061126709, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8719222545623779, "num_tokens": 365027050.0, "step": 9567 }, { "epoch": 1.2171479455540009, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.953721046447754, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8675215244293213, "num_tokens": 365060050.0, "step": 9568 }, { "epoch": 1.2172751558325912, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.88960075378418, "learning_rate": 1e-06, "loss": 0.503, "mean_token_accuracy": 0.8369736671447754, "num_tokens": 365098030.0, "step": 9569 }, { "epoch": 1.2174023661111817, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.930325508117676, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8701075315475464, "num_tokens": 365141143.0, "step": 9570 }, { "epoch": 1.2175295763897722, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.958378791809082, "learning_rate": 1e-06, "loss": 0.4985, "mean_token_accuracy": 0.8456594347953796, "num_tokens": 365174204.0, "step": 9571 }, { "epoch": 1.2176567866683627, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.927810668945312, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8686297535896301, "num_tokens": 365207170.0, "step": 9572 }, { "epoch": 1.2177839969469533, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.88017749786377, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8707383871078491, "num_tokens": 365248926.0, "step": 9573 }, { "epoch": 1.2179112072255438, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.923140525817871, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8711881041526794, "num_tokens": 365283588.0, "step": 9574 }, { "epoch": 1.2180384175041343, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.880742073059082, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8654001951217651, "num_tokens": 365324869.0, "step": 9575 }, { "epoch": 1.2181656277827249, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.873090744018555, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8637498617172241, "num_tokens": 365366832.0, "step": 9576 }, { "epoch": 1.2182928380613154, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.902409553527832, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8654145002365112, "num_tokens": 365401148.0, "step": 9577 }, { "epoch": 1.218420048339906, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.928130149841309, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8694819211959839, "num_tokens": 365442303.0, "step": 9578 }, { "epoch": 1.2185472586184964, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.843461990356445, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8809747695922852, "num_tokens": 365477877.0, "step": 9579 }, { "epoch": 1.218674468897087, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.916289329528809, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8565192818641663, "num_tokens": 365520648.0, "step": 9580 }, { "epoch": 1.2188016791756775, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.95749282836914, "learning_rate": 1e-06, "loss": 0.4761, "mean_token_accuracy": 0.8500915169715881, "num_tokens": 365555129.0, "step": 9581 }, { "epoch": 1.218928889454268, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.94363784790039, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8734912872314453, "num_tokens": 365594368.0, "step": 9582 }, { "epoch": 1.2190560997328583, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.937829971313477, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8678632974624634, "num_tokens": 365634763.0, "step": 9583 }, { "epoch": 1.2191833100114489, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.95096492767334, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8840334415435791, "num_tokens": 365675951.0, "step": 9584 }, { "epoch": 1.2193105202900394, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.87444019317627, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8755329251289368, "num_tokens": 365710348.0, "step": 9585 }, { "epoch": 1.21943773056863, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.00267219543457, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8705979585647583, "num_tokens": 365745166.0, "step": 9586 }, { "epoch": 1.2195649408472204, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.940144538879395, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8810404539108276, "num_tokens": 365780122.0, "step": 9587 }, { "epoch": 1.219692151125811, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.86116886138916, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8674416542053223, "num_tokens": 365816326.0, "step": 9588 }, { "epoch": 1.2198193614044015, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.976935386657715, "learning_rate": 1e-06, "loss": 0.4889, "mean_token_accuracy": 0.8422667980194092, "num_tokens": 365858938.0, "step": 9589 }, { "epoch": 1.219946571682992, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.99157428741455, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8578581809997559, "num_tokens": 365893892.0, "step": 9590 }, { "epoch": 1.2200737819615826, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.876791000366211, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8590726256370544, "num_tokens": 365938787.0, "step": 9591 }, { "epoch": 1.220200992240173, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.997114181518555, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8670863509178162, "num_tokens": 365972664.0, "step": 9592 }, { "epoch": 1.2203282025187634, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.890939712524414, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8740583062171936, "num_tokens": 366013367.0, "step": 9593 }, { "epoch": 1.220455412797354, "ewc_loss": 0.0263671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.955808639526367, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8739913105964661, "num_tokens": 366051637.0, "step": 9594 }, { "epoch": 1.2205826230759445, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.911920547485352, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8648074865341187, "num_tokens": 366092713.0, "step": 9595 }, { "epoch": 1.220709833354535, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.913525581359863, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8682760000228882, "num_tokens": 366131594.0, "step": 9596 }, { "epoch": 1.2208370436331255, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.921895027160645, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8623260855674744, "num_tokens": 366170871.0, "step": 9597 }, { "epoch": 1.220964253911716, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.941851615905762, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8573365211486816, "num_tokens": 366212974.0, "step": 9598 }, { "epoch": 1.2210914641903066, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.891222953796387, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8600062131881714, "num_tokens": 366255904.0, "step": 9599 }, { "epoch": 1.221218674468897, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.884746551513672, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8725290298461914, "num_tokens": 366288887.0, "step": 9600 }, { "epoch": 1.2213458847474876, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.909186363220215, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8806573748588562, "num_tokens": 366328796.0, "step": 9601 }, { "epoch": 1.2214730950260781, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.886578559875488, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8648846745491028, "num_tokens": 366373093.0, "step": 9602 }, { "epoch": 1.2216003053046687, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.868380546569824, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.876436173915863, "num_tokens": 366414353.0, "step": 9603 }, { "epoch": 1.2217275155832592, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.966813087463379, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8804352879524231, "num_tokens": 366450683.0, "step": 9604 }, { "epoch": 1.2218547258618497, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.937729835510254, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8730834722518921, "num_tokens": 366487158.0, "step": 9605 }, { "epoch": 1.2219819361404403, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.936134338378906, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8695510625839233, "num_tokens": 366529984.0, "step": 9606 }, { "epoch": 1.2221091464190306, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.942811965942383, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8709828853607178, "num_tokens": 366564100.0, "step": 9607 }, { "epoch": 1.222236356697621, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.934000015258789, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8753162026405334, "num_tokens": 366603752.0, "step": 9608 }, { "epoch": 1.2223635669762116, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.881779670715332, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8692790865898132, "num_tokens": 366642170.0, "step": 9609 }, { "epoch": 1.2224907772548022, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.93334674835205, "learning_rate": 1e-06, "loss": 0.4641, "mean_token_accuracy": 0.8545477390289307, "num_tokens": 366684038.0, "step": 9610 }, { "epoch": 1.2226179875333927, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.036247253417969, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.865293025970459, "num_tokens": 366717832.0, "step": 9611 }, { "epoch": 1.2227451978119832, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.830256462097168, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8623136281967163, "num_tokens": 366757999.0, "step": 9612 }, { "epoch": 1.2228724080905737, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.966047286987305, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8771263360977173, "num_tokens": 366798533.0, "step": 9613 }, { "epoch": 1.2229996183691643, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.953208923339844, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8649284839630127, "num_tokens": 366843066.0, "step": 9614 }, { "epoch": 1.2231268286477548, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.905989646911621, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8836069703102112, "num_tokens": 366883190.0, "step": 9615 }, { "epoch": 1.2232540389263453, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.98705768585205, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8721232414245605, "num_tokens": 366920642.0, "step": 9616 }, { "epoch": 1.2233812492049359, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.934588432312012, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8686689138412476, "num_tokens": 366954468.0, "step": 9617 }, { "epoch": 1.2235084594835262, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.955117225646973, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8531575202941895, "num_tokens": 366991170.0, "step": 9618 }, { "epoch": 1.2236356697621167, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.975024223327637, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8722547292709351, "num_tokens": 367031833.0, "step": 9619 }, { "epoch": 1.2237628800407072, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.84799861907959, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.86348557472229, "num_tokens": 367074709.0, "step": 9620 }, { "epoch": 1.2238900903192977, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.007204055786133, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8650844097137451, "num_tokens": 367106820.0, "step": 9621 }, { "epoch": 1.2240173005978883, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.914497375488281, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8812254667282104, "num_tokens": 367142298.0, "step": 9622 }, { "epoch": 1.2241445108764788, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.981365203857422, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.859154224395752, "num_tokens": 367179957.0, "step": 9623 }, { "epoch": 1.2242717211550693, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.897007942199707, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8699794411659241, "num_tokens": 367226800.0, "step": 9624 }, { "epoch": 1.2243989314336599, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.990161895751953, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8723753690719604, "num_tokens": 367257168.0, "step": 9625 }, { "epoch": 1.2245261417122504, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.88932991027832, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8700868487358093, "num_tokens": 367295254.0, "step": 9626 }, { "epoch": 1.224653351990841, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.941570281982422, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8558142781257629, "num_tokens": 367335814.0, "step": 9627 }, { "epoch": 1.2247805622694314, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.942794799804688, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8653082847595215, "num_tokens": 367377592.0, "step": 9628 }, { "epoch": 1.224907772548022, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.967988967895508, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8603401780128479, "num_tokens": 367408069.0, "step": 9629 }, { "epoch": 1.2250349828266125, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.914681434631348, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.868628978729248, "num_tokens": 367444787.0, "step": 9630 }, { "epoch": 1.225162193105203, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.924079895019531, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.854687511920929, "num_tokens": 367486848.0, "step": 9631 }, { "epoch": 1.2252894033837933, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.930505752563477, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8785830140113831, "num_tokens": 367530456.0, "step": 9632 }, { "epoch": 1.2254166136623839, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.929898262023926, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8713710904121399, "num_tokens": 367566513.0, "step": 9633 }, { "epoch": 1.2255438239409744, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.919358253479004, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.847762942314148, "num_tokens": 367607757.0, "step": 9634 }, { "epoch": 1.225671034219565, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.919382095336914, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8704826831817627, "num_tokens": 367645525.0, "step": 9635 }, { "epoch": 1.2257982444981554, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.996798515319824, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8764668703079224, "num_tokens": 367680840.0, "step": 9636 }, { "epoch": 1.225925454776746, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.869226455688477, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8758605718612671, "num_tokens": 367717444.0, "step": 9637 }, { "epoch": 1.2260526650553365, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.036577224731445, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8465479612350464, "num_tokens": 367756548.0, "step": 9638 }, { "epoch": 1.226179875333927, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.912820816040039, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8618719577789307, "num_tokens": 367798341.0, "step": 9639 }, { "epoch": 1.2263070856125176, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 12.004876136779785, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8745334148406982, "num_tokens": 367835087.0, "step": 9640 }, { "epoch": 1.226434295891108, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.95309066772461, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8696796298027039, "num_tokens": 367873816.0, "step": 9641 }, { "epoch": 1.2265615061696984, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.8831148147583, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8637720942497253, "num_tokens": 367913621.0, "step": 9642 }, { "epoch": 1.226688716448289, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.95852279663086, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8594311475753784, "num_tokens": 367946881.0, "step": 9643 }, { "epoch": 1.2268159267268794, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.88710880279541, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.871752142906189, "num_tokens": 367980806.0, "step": 9644 }, { "epoch": 1.22694313700547, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.946175575256348, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8751713037490845, "num_tokens": 368016041.0, "step": 9645 }, { "epoch": 1.2270703472840605, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.916694641113281, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8660138845443726, "num_tokens": 368052331.0, "step": 9646 }, { "epoch": 1.227197557562651, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.941250801086426, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8768128752708435, "num_tokens": 368091458.0, "step": 9647 }, { "epoch": 1.2273247678412416, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.966595649719238, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8563476800918579, "num_tokens": 368132433.0, "step": 9648 }, { "epoch": 1.227451978119832, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.90411376953125, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8685051202774048, "num_tokens": 368171030.0, "step": 9649 }, { "epoch": 1.2275791883984226, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.94962215423584, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8565070629119873, "num_tokens": 368211555.0, "step": 9650 }, { "epoch": 1.2277063986770131, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.950711250305176, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8742290735244751, "num_tokens": 368252076.0, "step": 9651 }, { "epoch": 1.2278336089556037, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.91305160522461, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8614115715026855, "num_tokens": 368291265.0, "step": 9652 }, { "epoch": 1.2279608192341942, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.952502250671387, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8608129620552063, "num_tokens": 368332178.0, "step": 9653 }, { "epoch": 1.2280880295127847, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.900259017944336, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8656284213066101, "num_tokens": 368374103.0, "step": 9654 }, { "epoch": 1.2282152397913753, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.963363647460938, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8706676363945007, "num_tokens": 368411928.0, "step": 9655 }, { "epoch": 1.2283424500699656, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.909367561340332, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8442461490631104, "num_tokens": 368449200.0, "step": 9656 }, { "epoch": 1.228469660348556, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.917017936706543, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8504420518875122, "num_tokens": 368489935.0, "step": 9657 }, { "epoch": 1.2285968706271466, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.958885192871094, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8815768957138062, "num_tokens": 368522227.0, "step": 9658 }, { "epoch": 1.2287240809057371, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.927385330200195, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8666958808898926, "num_tokens": 368555084.0, "step": 9659 }, { "epoch": 1.2288512911843277, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.963071823120117, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8633849620819092, "num_tokens": 368591845.0, "step": 9660 }, { "epoch": 1.2289785014629182, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.88181209564209, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8569296598434448, "num_tokens": 368637841.0, "step": 9661 }, { "epoch": 1.2291057117415087, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.911612510681152, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8828797936439514, "num_tokens": 368677470.0, "step": 9662 }, { "epoch": 1.2292329220200993, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.912158012390137, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8815363645553589, "num_tokens": 368718412.0, "step": 9663 }, { "epoch": 1.2293601322986898, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.923951148986816, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.856220006942749, "num_tokens": 368759267.0, "step": 9664 }, { "epoch": 1.2294873425772803, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.95915699005127, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8592209815979004, "num_tokens": 368797586.0, "step": 9665 }, { "epoch": 1.2296145528558708, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.916895866394043, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8600301742553711, "num_tokens": 368833801.0, "step": 9666 }, { "epoch": 1.2297417631344612, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.940589904785156, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8512952327728271, "num_tokens": 368878513.0, "step": 9667 }, { "epoch": 1.2298689734130517, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.950370788574219, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8603581190109253, "num_tokens": 368917815.0, "step": 9668 }, { "epoch": 1.2299961836916422, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.949130058288574, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.856399655342102, "num_tokens": 368954971.0, "step": 9669 }, { "epoch": 1.2301233939702327, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.97175407409668, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8726701736450195, "num_tokens": 368996212.0, "step": 9670 }, { "epoch": 1.2302506042488233, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.929966926574707, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8808614015579224, "num_tokens": 369032186.0, "step": 9671 }, { "epoch": 1.2303778145274138, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.973995208740234, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8713988661766052, "num_tokens": 369077565.0, "step": 9672 }, { "epoch": 1.2305050248060043, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.947458267211914, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.873850405216217, "num_tokens": 369111935.0, "step": 9673 }, { "epoch": 1.2306322350845948, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.981425285339355, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.865531325340271, "num_tokens": 369154374.0, "step": 9674 }, { "epoch": 1.2307594453631854, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.917455673217773, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8590825796127319, "num_tokens": 369198557.0, "step": 9675 }, { "epoch": 1.230886655641776, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.968791007995605, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8510290384292603, "num_tokens": 369237756.0, "step": 9676 }, { "epoch": 1.2310138659203664, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.952775001525879, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.85467129945755, "num_tokens": 369270155.0, "step": 9677 }, { "epoch": 1.231141076198957, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.946333885192871, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8597486615180969, "num_tokens": 369310870.0, "step": 9678 }, { "epoch": 1.2312682864775475, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.973841667175293, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8496658802032471, "num_tokens": 369344679.0, "step": 9679 }, { "epoch": 1.231395496756138, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.960739135742188, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8670228719711304, "num_tokens": 369380388.0, "step": 9680 }, { "epoch": 1.2315227070347283, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.923870086669922, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.878600001335144, "num_tokens": 369422327.0, "step": 9681 }, { "epoch": 1.2316499173133189, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.964279174804688, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8736073970794678, "num_tokens": 369457160.0, "step": 9682 }, { "epoch": 1.2317771275919094, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.939835548400879, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8670676350593567, "num_tokens": 369497883.0, "step": 9683 }, { "epoch": 1.2319043378705, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.936654090881348, "learning_rate": 1e-06, "loss": 0.324, "mean_token_accuracy": 0.8965681195259094, "num_tokens": 369540014.0, "step": 9684 }, { "epoch": 1.2320315481490904, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.913985252380371, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8657686710357666, "num_tokens": 369580424.0, "step": 9685 }, { "epoch": 1.232158758427681, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.966312408447266, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8633336424827576, "num_tokens": 369623615.0, "step": 9686 }, { "epoch": 1.2322859687062715, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.907266616821289, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8559038639068604, "num_tokens": 369658408.0, "step": 9687 }, { "epoch": 1.232413178984862, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.952313423156738, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8640780448913574, "num_tokens": 369696311.0, "step": 9688 }, { "epoch": 1.2325403892634526, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.886578559875488, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8545341491699219, "num_tokens": 369737473.0, "step": 9689 }, { "epoch": 1.232667599542043, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.952001571655273, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.864396333694458, "num_tokens": 369770151.0, "step": 9690 }, { "epoch": 1.2327948098206334, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 11.93690013885498, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8856620192527771, "num_tokens": 369813503.0, "step": 9691 }, { "epoch": 1.232922020099224, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.902189254760742, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8678368330001831, "num_tokens": 369848768.0, "step": 9692 }, { "epoch": 1.2330492303778144, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.976510047912598, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8698320388793945, "num_tokens": 369892510.0, "step": 9693 }, { "epoch": 1.233176440656405, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.931394577026367, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8737778663635254, "num_tokens": 369929638.0, "step": 9694 }, { "epoch": 1.2333036509349955, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.918014526367188, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8876837491989136, "num_tokens": 369971201.0, "step": 9695 }, { "epoch": 1.233430861213586, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.920403480529785, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8613859415054321, "num_tokens": 370011713.0, "step": 9696 }, { "epoch": 1.2335580714921766, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.970943450927734, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8477005958557129, "num_tokens": 370042647.0, "step": 9697 }, { "epoch": 1.233685281770767, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.910892486572266, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8653955459594727, "num_tokens": 370083976.0, "step": 9698 }, { "epoch": 1.2338124920493576, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.934059143066406, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.863762378692627, "num_tokens": 370126435.0, "step": 9699 }, { "epoch": 1.2339397023279481, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.99654769897461, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.876484215259552, "num_tokens": 370165038.0, "step": 9700 }, { "epoch": 1.2340669126065387, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.938260078430176, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8588292598724365, "num_tokens": 370203438.0, "step": 9701 }, { "epoch": 1.2341941228851292, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.940828323364258, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.882191002368927, "num_tokens": 370235144.0, "step": 9702 }, { "epoch": 1.2343213331637197, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.90395450592041, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8618788719177246, "num_tokens": 370278936.0, "step": 9703 }, { "epoch": 1.2344485434423103, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.915870666503906, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8715182542800903, "num_tokens": 370312595.0, "step": 9704 }, { "epoch": 1.2345757537209006, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.958333969116211, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8642339110374451, "num_tokens": 370351697.0, "step": 9705 }, { "epoch": 1.234702963999491, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.935946464538574, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8801222443580627, "num_tokens": 370391264.0, "step": 9706 }, { "epoch": 1.2348301742780816, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.953573226928711, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8741467595100403, "num_tokens": 370425248.0, "step": 9707 }, { "epoch": 1.2349573845566721, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.927724838256836, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.8458859920501709, "num_tokens": 370462670.0, "step": 9708 }, { "epoch": 1.2350845948352627, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.918277740478516, "learning_rate": 1e-06, "loss": 0.4746, "mean_token_accuracy": 0.8466139435768127, "num_tokens": 370506126.0, "step": 9709 }, { "epoch": 1.2352118051138532, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.902650833129883, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8774943351745605, "num_tokens": 370546895.0, "step": 9710 }, { "epoch": 1.2353390153924437, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.969735145568848, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8715251684188843, "num_tokens": 370590645.0, "step": 9711 }, { "epoch": 1.2354662256710343, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.93526554107666, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8700211048126221, "num_tokens": 370628614.0, "step": 9712 }, { "epoch": 1.2355934359496248, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.965372085571289, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8693624138832092, "num_tokens": 370666082.0, "step": 9713 }, { "epoch": 1.2357206462282153, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.939175605773926, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8530539274215698, "num_tokens": 370705780.0, "step": 9714 }, { "epoch": 1.2358478565068058, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.97071361541748, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8722094297409058, "num_tokens": 370742550.0, "step": 9715 }, { "epoch": 1.2359750667853961, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.936847686767578, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.871540367603302, "num_tokens": 370784406.0, "step": 9716 }, { "epoch": 1.2361022770639867, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.997262954711914, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8623309135437012, "num_tokens": 370820911.0, "step": 9717 }, { "epoch": 1.2362294873425772, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.998360633850098, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.878980278968811, "num_tokens": 370857427.0, "step": 9718 }, { "epoch": 1.2363566976211677, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.925061225891113, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8574643135070801, "num_tokens": 370897989.0, "step": 9719 }, { "epoch": 1.2364839078997583, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.029284477233887, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8544401526451111, "num_tokens": 370932672.0, "step": 9720 }, { "epoch": 1.2366111181783488, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.943939208984375, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.858635663986206, "num_tokens": 370971027.0, "step": 9721 }, { "epoch": 1.2367383284569393, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.958474159240723, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.877509593963623, "num_tokens": 371014742.0, "step": 9722 }, { "epoch": 1.2368655387355298, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.947086334228516, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.850459098815918, "num_tokens": 371052065.0, "step": 9723 }, { "epoch": 1.2369927490141204, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.995810508728027, "learning_rate": 1e-06, "loss": 0.4665, "mean_token_accuracy": 0.8457391858100891, "num_tokens": 371093897.0, "step": 9724 }, { "epoch": 1.237119959292711, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.948809623718262, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8668010234832764, "num_tokens": 371132390.0, "step": 9725 }, { "epoch": 1.2372471695713014, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.966926574707031, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8787988424301147, "num_tokens": 371170468.0, "step": 9726 }, { "epoch": 1.237374379849892, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.920085906982422, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8639711141586304, "num_tokens": 371209323.0, "step": 9727 }, { "epoch": 1.2375015901284825, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.933176040649414, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8674010038375854, "num_tokens": 371246526.0, "step": 9728 }, { "epoch": 1.237628800407073, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.925963401794434, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8706337213516235, "num_tokens": 371282850.0, "step": 9729 }, { "epoch": 1.2377560106856633, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.922550201416016, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.875873327255249, "num_tokens": 371312614.0, "step": 9730 }, { "epoch": 1.2378832209642538, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.938772201538086, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8699766397476196, "num_tokens": 371350170.0, "step": 9731 }, { "epoch": 1.2380104312428444, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.909751892089844, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8737751245498657, "num_tokens": 371389295.0, "step": 9732 }, { "epoch": 1.238137641521435, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.925737380981445, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8676461577415466, "num_tokens": 371430579.0, "step": 9733 }, { "epoch": 1.2382648518000254, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.958166122436523, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.857768177986145, "num_tokens": 371465087.0, "step": 9734 }, { "epoch": 1.238392062078616, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.955350875854492, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8747174739837646, "num_tokens": 371503549.0, "step": 9735 }, { "epoch": 1.2385192723572065, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.93710994720459, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8868796825408936, "num_tokens": 371542953.0, "step": 9736 }, { "epoch": 1.238646482635797, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.893970489501953, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8714781999588013, "num_tokens": 371582541.0, "step": 9737 }, { "epoch": 1.2387736929143875, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.023575782775879, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8772920370101929, "num_tokens": 371617221.0, "step": 9738 }, { "epoch": 1.238900903192978, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.973901748657227, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8707237243652344, "num_tokens": 371660773.0, "step": 9739 }, { "epoch": 1.2390281134715684, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.985981941223145, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8626548647880554, "num_tokens": 371694763.0, "step": 9740 }, { "epoch": 1.239155323750159, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.95357608795166, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8670178055763245, "num_tokens": 371734793.0, "step": 9741 }, { "epoch": 1.2392825340287494, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.988417625427246, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8624582290649414, "num_tokens": 371775691.0, "step": 9742 }, { "epoch": 1.23940974430734, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.980213165283203, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8591929078102112, "num_tokens": 371817319.0, "step": 9743 }, { "epoch": 1.2395369545859305, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.001996994018555, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8669285178184509, "num_tokens": 371847431.0, "step": 9744 }, { "epoch": 1.239664164864521, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.912671089172363, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8808733224868774, "num_tokens": 371881788.0, "step": 9745 }, { "epoch": 1.2397913751431116, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.992958068847656, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8630122542381287, "num_tokens": 371917388.0, "step": 9746 }, { "epoch": 1.239918585421702, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.960076332092285, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8762294054031372, "num_tokens": 371957561.0, "step": 9747 }, { "epoch": 1.2400457957002926, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.041011810302734, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8610525131225586, "num_tokens": 371987322.0, "step": 9748 }, { "epoch": 1.2401730059788831, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.938436508178711, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8702789545059204, "num_tokens": 372027263.0, "step": 9749 }, { "epoch": 1.2403002162574737, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.946998596191406, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8716944456100464, "num_tokens": 372061441.0, "step": 9750 }, { "epoch": 1.2404274265360642, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.927420616149902, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8725239038467407, "num_tokens": 372099085.0, "step": 9751 }, { "epoch": 1.2405546368146547, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.97732925415039, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.869558572769165, "num_tokens": 372134106.0, "step": 9752 }, { "epoch": 1.2406818470932452, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.977731704711914, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8764750361442566, "num_tokens": 372164635.0, "step": 9753 }, { "epoch": 1.2408090573718356, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.00964069366455, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8736121654510498, "num_tokens": 372204924.0, "step": 9754 }, { "epoch": 1.240936267650426, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.942792892456055, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8672749996185303, "num_tokens": 372240939.0, "step": 9755 }, { "epoch": 1.2410634779290166, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.011924743652344, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8628804683685303, "num_tokens": 372272355.0, "step": 9756 }, { "epoch": 1.2411906882076071, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.945313453674316, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8655292987823486, "num_tokens": 372309566.0, "step": 9757 }, { "epoch": 1.2413178984861977, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.007800102233887, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8624218702316284, "num_tokens": 372347768.0, "step": 9758 }, { "epoch": 1.2414451087647882, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.924454689025879, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8763716816902161, "num_tokens": 372381020.0, "step": 9759 }, { "epoch": 1.2415723190433787, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.992923736572266, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.85666823387146, "num_tokens": 372422434.0, "step": 9760 }, { "epoch": 1.2416995293219693, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.932748794555664, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.858692467212677, "num_tokens": 372459342.0, "step": 9761 }, { "epoch": 1.2418267396005598, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.958494186401367, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8705264925956726, "num_tokens": 372496637.0, "step": 9762 }, { "epoch": 1.2419539498791503, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.897604942321777, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8802831172943115, "num_tokens": 372532068.0, "step": 9763 }, { "epoch": 1.2420811601577408, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.020034790039062, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8742966651916504, "num_tokens": 372565460.0, "step": 9764 }, { "epoch": 1.2422083704363311, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.941534996032715, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8713904619216919, "num_tokens": 372605998.0, "step": 9765 }, { "epoch": 1.2423355807149217, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.91988754272461, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8569164872169495, "num_tokens": 372642038.0, "step": 9766 }, { "epoch": 1.2424627909935122, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.915589332580566, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.864296555519104, "num_tokens": 372678192.0, "step": 9767 }, { "epoch": 1.2425900012721027, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.98382568359375, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8643441796302795, "num_tokens": 372712867.0, "step": 9768 }, { "epoch": 1.2427172115506933, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.059455871582031, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8508387804031372, "num_tokens": 372745790.0, "step": 9769 }, { "epoch": 1.2428444218292838, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.937480926513672, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8781911134719849, "num_tokens": 372777634.0, "step": 9770 }, { "epoch": 1.2429716321078743, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.973731994628906, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8528128862380981, "num_tokens": 372820527.0, "step": 9771 }, { "epoch": 1.2430988423864648, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.947855949401855, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8776213526725769, "num_tokens": 372858479.0, "step": 9772 }, { "epoch": 1.2432260526650554, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.987762451171875, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8745154142379761, "num_tokens": 372895219.0, "step": 9773 }, { "epoch": 1.243353262943646, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.921406745910645, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8797215223312378, "num_tokens": 372934491.0, "step": 9774 }, { "epoch": 1.2434804732222364, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.949700355529785, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.858761191368103, "num_tokens": 372972892.0, "step": 9775 }, { "epoch": 1.243607683500827, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.997611999511719, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8705819845199585, "num_tokens": 373011564.0, "step": 9776 }, { "epoch": 1.2437348937794175, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.030349731445312, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8691536784172058, "num_tokens": 373049444.0, "step": 9777 }, { "epoch": 1.243862104058008, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.96291732788086, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8675857782363892, "num_tokens": 373085466.0, "step": 9778 }, { "epoch": 1.2439893143365983, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.96511459350586, "learning_rate": 1e-06, "loss": 0.475, "mean_token_accuracy": 0.8454040288925171, "num_tokens": 373124712.0, "step": 9779 }, { "epoch": 1.2441165246151888, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.017687797546387, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8733845353126526, "num_tokens": 373160008.0, "step": 9780 }, { "epoch": 1.2442437348937794, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.013422966003418, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8685175180435181, "num_tokens": 373192110.0, "step": 9781 }, { "epoch": 1.24437094517237, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.001253128051758, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8697782754898071, "num_tokens": 373234432.0, "step": 9782 }, { "epoch": 1.2444981554509604, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.920467376708984, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8584132194519043, "num_tokens": 373270725.0, "step": 9783 }, { "epoch": 1.244625365729551, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.045858383178711, "learning_rate": 1e-06, "loss": 0.4942, "mean_token_accuracy": 0.8462429642677307, "num_tokens": 373307531.0, "step": 9784 }, { "epoch": 1.2447525760081415, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.9406156539917, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.871425449848175, "num_tokens": 373351657.0, "step": 9785 }, { "epoch": 1.244879786286732, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.984626770019531, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8729819059371948, "num_tokens": 373388811.0, "step": 9786 }, { "epoch": 1.2450069965653225, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.933446884155273, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8773643970489502, "num_tokens": 373423378.0, "step": 9787 }, { "epoch": 1.245134206843913, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.99988079071045, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8687653541564941, "num_tokens": 373460982.0, "step": 9788 }, { "epoch": 1.2452614171225034, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.016915321350098, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.874123215675354, "num_tokens": 373501474.0, "step": 9789 }, { "epoch": 1.245388627401094, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.952021598815918, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8719979524612427, "num_tokens": 373537830.0, "step": 9790 }, { "epoch": 1.2455158376796844, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.986749649047852, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8605105876922607, "num_tokens": 373573975.0, "step": 9791 }, { "epoch": 1.245643047958275, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.980603218078613, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8473950624465942, "num_tokens": 373612157.0, "step": 9792 }, { "epoch": 1.2457702582368655, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.9775390625, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8874064087867737, "num_tokens": 373645065.0, "step": 9793 }, { "epoch": 1.245897468515456, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.976865768432617, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8626093864440918, "num_tokens": 373687658.0, "step": 9794 }, { "epoch": 1.2460246787940465, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.93203353881836, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8689424991607666, "num_tokens": 373732262.0, "step": 9795 }, { "epoch": 1.246151889072637, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.973727226257324, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8486193418502808, "num_tokens": 373772410.0, "step": 9796 }, { "epoch": 1.2462790993512276, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.96606731414795, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8686338067054749, "num_tokens": 373808773.0, "step": 9797 }, { "epoch": 1.2464063096298181, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.9644193649292, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8629539608955383, "num_tokens": 373848945.0, "step": 9798 }, { "epoch": 1.2465335199084087, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.944254875183105, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8629663586616516, "num_tokens": 373892716.0, "step": 9799 }, { "epoch": 1.2466607301869992, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.904609680175781, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8831886053085327, "num_tokens": 373929443.0, "step": 9800 }, { "epoch": 1.2467879404655897, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.947430610656738, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.868523895740509, "num_tokens": 373968882.0, "step": 9801 }, { "epoch": 1.2469151507441802, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.985987663269043, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8800621032714844, "num_tokens": 374009265.0, "step": 9802 }, { "epoch": 1.2470423610227706, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.960868835449219, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8529932498931885, "num_tokens": 374047152.0, "step": 9803 }, { "epoch": 1.247169571301361, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.985947608947754, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8643747568130493, "num_tokens": 374083127.0, "step": 9804 }, { "epoch": 1.2472967815799516, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.945703506469727, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8763861060142517, "num_tokens": 374127086.0, "step": 9805 }, { "epoch": 1.2474239918585421, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.98283863067627, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8617736101150513, "num_tokens": 374163879.0, "step": 9806 }, { "epoch": 1.2475512021371327, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.97386360168457, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8724206686019897, "num_tokens": 374206639.0, "step": 9807 }, { "epoch": 1.2476784124157232, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.964485168457031, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8646441698074341, "num_tokens": 374244106.0, "step": 9808 }, { "epoch": 1.2478056226943137, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.93501091003418, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8637611269950867, "num_tokens": 374285776.0, "step": 9809 }, { "epoch": 1.2479328329729042, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.969877243041992, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8767248392105103, "num_tokens": 374323036.0, "step": 9810 }, { "epoch": 1.2480600432514948, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.903380393981934, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8743136525154114, "num_tokens": 374360204.0, "step": 9811 }, { "epoch": 1.2481872535300853, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.01574420928955, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8773682713508606, "num_tokens": 374394961.0, "step": 9812 }, { "epoch": 1.2483144638086758, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.940733909606934, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8691139221191406, "num_tokens": 374436142.0, "step": 9813 }, { "epoch": 1.2484416740872661, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.019298553466797, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8647363185882568, "num_tokens": 374474263.0, "step": 9814 }, { "epoch": 1.2485688843658567, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.9600248336792, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8620598316192627, "num_tokens": 374510250.0, "step": 9815 }, { "epoch": 1.2486960946444472, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.919002532958984, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8857442140579224, "num_tokens": 374549579.0, "step": 9816 }, { "epoch": 1.2488233049230377, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.01370620727539, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8660876750946045, "num_tokens": 374588631.0, "step": 9817 }, { "epoch": 1.2489505152016283, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.933503150939941, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8632931709289551, "num_tokens": 374627092.0, "step": 9818 }, { "epoch": 1.2490777254802188, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.000846862792969, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8735700845718384, "num_tokens": 374657797.0, "step": 9819 }, { "epoch": 1.2492049357588093, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.942148208618164, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8718770146369934, "num_tokens": 374694011.0, "step": 9820 }, { "epoch": 1.2493321460373998, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.88078498840332, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8695442080497742, "num_tokens": 374735127.0, "step": 9821 }, { "epoch": 1.2494593563159904, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.961003303527832, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8648275136947632, "num_tokens": 374774170.0, "step": 9822 }, { "epoch": 1.249586566594581, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.995810508728027, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.868058443069458, "num_tokens": 374809162.0, "step": 9823 }, { "epoch": 1.2497137768731714, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.932374954223633, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8599671125411987, "num_tokens": 374847974.0, "step": 9824 }, { "epoch": 1.249840987151762, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.944445610046387, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8837920427322388, "num_tokens": 374894911.0, "step": 9825 }, { "epoch": 1.2499681974303525, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.998360633850098, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8629599809646606, "num_tokens": 374933534.0, "step": 9826 }, { "epoch": 1.250095407708943, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.99029541015625, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8558570146560669, "num_tokens": 374968183.0, "step": 9827 }, { "epoch": 1.2502226179875333, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.951573371887207, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8612210750579834, "num_tokens": 375005973.0, "step": 9828 }, { "epoch": 1.2503498282661238, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 11.985441207885742, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8611978888511658, "num_tokens": 375043446.0, "step": 9829 }, { "epoch": 1.2504770385447144, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.958296775817871, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.87147057056427, "num_tokens": 375081518.0, "step": 9830 }, { "epoch": 1.250604248823305, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.993870735168457, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8608819246292114, "num_tokens": 375117767.0, "step": 9831 }, { "epoch": 1.2507314591018954, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.012036323547363, "learning_rate": 1e-06, "loss": 0.4848, "mean_token_accuracy": 0.8435543775558472, "num_tokens": 375152061.0, "step": 9832 }, { "epoch": 1.250858669380486, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.970351219177246, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8614184856414795, "num_tokens": 375195677.0, "step": 9833 }, { "epoch": 1.2509858796590765, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.959443092346191, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8611898422241211, "num_tokens": 375230319.0, "step": 9834 }, { "epoch": 1.251113089937667, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.993078231811523, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8720247745513916, "num_tokens": 375268970.0, "step": 9835 }, { "epoch": 1.2512403002162575, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.975672721862793, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8739791512489319, "num_tokens": 375303661.0, "step": 9836 }, { "epoch": 1.2513675104948478, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.024227142333984, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8747097253799438, "num_tokens": 375343278.0, "step": 9837 }, { "epoch": 1.2514947207734384, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.989978790283203, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8744462728500366, "num_tokens": 375378574.0, "step": 9838 }, { "epoch": 1.251621931052029, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.997990608215332, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.863205075263977, "num_tokens": 375412781.0, "step": 9839 }, { "epoch": 1.2517491413306194, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.994915008544922, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8798291683197021, "num_tokens": 375450431.0, "step": 9840 }, { "epoch": 1.25187635160921, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.921211242675781, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8624241352081299, "num_tokens": 375490692.0, "step": 9841 }, { "epoch": 1.2520035618878005, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.03781795501709, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8732860088348389, "num_tokens": 375535278.0, "step": 9842 }, { "epoch": 1.252130772166391, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.014138221740723, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8613007068634033, "num_tokens": 375575859.0, "step": 9843 }, { "epoch": 1.2522579824449815, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.020976066589355, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8607054948806763, "num_tokens": 375611084.0, "step": 9844 }, { "epoch": 1.252385192723572, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.971662521362305, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8783107995986938, "num_tokens": 375644926.0, "step": 9845 }, { "epoch": 1.2525124030021626, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.025788307189941, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8761447668075562, "num_tokens": 375686970.0, "step": 9846 }, { "epoch": 1.2526396132807531, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.98302936553955, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8789763450622559, "num_tokens": 375727372.0, "step": 9847 }, { "epoch": 1.2527668235593437, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.009834289550781, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.856643557548523, "num_tokens": 375763568.0, "step": 9848 }, { "epoch": 1.2528940338379342, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.038738250732422, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8507905006408691, "num_tokens": 375797954.0, "step": 9849 }, { "epoch": 1.2530212441165247, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.966818809509277, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.866269052028656, "num_tokens": 375836362.0, "step": 9850 }, { "epoch": 1.2531484543951152, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.96084213256836, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8630853891372681, "num_tokens": 375879403.0, "step": 9851 }, { "epoch": 1.2532756646737058, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.021888732910156, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8704842329025269, "num_tokens": 375916728.0, "step": 9852 }, { "epoch": 1.253402874952296, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.997297286987305, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8616777658462524, "num_tokens": 375951743.0, "step": 9853 }, { "epoch": 1.2535300852308866, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.98698616027832, "learning_rate": 1e-06, "loss": 0.4897, "mean_token_accuracy": 0.8425900936126709, "num_tokens": 375988872.0, "step": 9854 }, { "epoch": 1.2536572955094771, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.97044563293457, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.8568849563598633, "num_tokens": 376030388.0, "step": 9855 }, { "epoch": 1.2537845057880677, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.00730037689209, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8758642673492432, "num_tokens": 376062975.0, "step": 9856 }, { "epoch": 1.2539117160666582, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.975166320800781, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8604788780212402, "num_tokens": 376101873.0, "step": 9857 }, { "epoch": 1.2540389263452487, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.006852149963379, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8476397395133972, "num_tokens": 376138774.0, "step": 9858 }, { "epoch": 1.2541661366238392, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.024928092956543, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8595293760299683, "num_tokens": 376172357.0, "step": 9859 }, { "epoch": 1.2542933469024298, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.99095630645752, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8574904203414917, "num_tokens": 376212977.0, "step": 9860 }, { "epoch": 1.2544205571810203, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.964261054992676, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8570938110351562, "num_tokens": 376256989.0, "step": 9861 }, { "epoch": 1.2545477674596106, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.994518280029297, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8718271255493164, "num_tokens": 376300130.0, "step": 9862 }, { "epoch": 1.2546749777382011, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.953190803527832, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8701428174972534, "num_tokens": 376337475.0, "step": 9863 }, { "epoch": 1.2548021880167917, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.025052070617676, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8588402271270752, "num_tokens": 376376485.0, "step": 9864 }, { "epoch": 1.2549293982953822, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.942422866821289, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8728395700454712, "num_tokens": 376413393.0, "step": 9865 }, { "epoch": 1.2550566085739727, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.979289054870605, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8571842908859253, "num_tokens": 376448192.0, "step": 9866 }, { "epoch": 1.2551838188525632, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.941106796264648, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8696755170822144, "num_tokens": 376489382.0, "step": 9867 }, { "epoch": 1.2553110291311538, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.016877174377441, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8711955547332764, "num_tokens": 376527221.0, "step": 9868 }, { "epoch": 1.2554382394097443, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.91678524017334, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8631948232650757, "num_tokens": 376567045.0, "step": 9869 }, { "epoch": 1.2555654496883348, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.981959342956543, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8736766576766968, "num_tokens": 376607039.0, "step": 9870 }, { "epoch": 1.2556926599669254, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.977546691894531, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8699066638946533, "num_tokens": 376643220.0, "step": 9871 }, { "epoch": 1.255819870245516, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.973312377929688, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.886573076248169, "num_tokens": 376681417.0, "step": 9872 }, { "epoch": 1.2559470805241064, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.976025581359863, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8851259350776672, "num_tokens": 376722836.0, "step": 9873 }, { "epoch": 1.256074290802697, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.961912155151367, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8602039813995361, "num_tokens": 376763194.0, "step": 9874 }, { "epoch": 1.2562015010812875, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.000786781311035, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8623309135437012, "num_tokens": 376803701.0, "step": 9875 }, { "epoch": 1.256328711359878, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.962387084960938, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8763962984085083, "num_tokens": 376835296.0, "step": 9876 }, { "epoch": 1.2564559216384683, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.961669921875, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8710391521453857, "num_tokens": 376876317.0, "step": 9877 }, { "epoch": 1.2565831319170588, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.965805053710938, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8658561110496521, "num_tokens": 376916403.0, "step": 9878 }, { "epoch": 1.2567103421956494, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.951878547668457, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8759817481040955, "num_tokens": 376956079.0, "step": 9879 }, { "epoch": 1.25683755247424, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.007205963134766, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8575528860092163, "num_tokens": 376991618.0, "step": 9880 }, { "epoch": 1.2569647627528304, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.032522201538086, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8740038275718689, "num_tokens": 377033964.0, "step": 9881 }, { "epoch": 1.257091973031421, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.97851848602295, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8746252655982971, "num_tokens": 377070091.0, "step": 9882 }, { "epoch": 1.2572191833100115, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.031047821044922, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8756493330001831, "num_tokens": 377111185.0, "step": 9883 }, { "epoch": 1.257346393588602, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.95351791381836, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8721445798873901, "num_tokens": 377147637.0, "step": 9884 }, { "epoch": 1.2574736038671925, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.062259674072266, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8665894865989685, "num_tokens": 377181918.0, "step": 9885 }, { "epoch": 1.2576008141457828, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.982173919677734, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8616805672645569, "num_tokens": 377222900.0, "step": 9886 }, { "epoch": 1.2577280244243734, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.980792045593262, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8815046548843384, "num_tokens": 377263107.0, "step": 9887 }, { "epoch": 1.257855234702964, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.011241912841797, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8832106590270996, "num_tokens": 377301384.0, "step": 9888 }, { "epoch": 1.2579824449815544, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.983688354492188, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8608145117759705, "num_tokens": 377339865.0, "step": 9889 }, { "epoch": 1.258109655260145, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.007283210754395, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8761875033378601, "num_tokens": 377381582.0, "step": 9890 }, { "epoch": 1.2582368655387355, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.98541259765625, "learning_rate": 1e-06, "loss": 0.4844, "mean_token_accuracy": 0.8452759981155396, "num_tokens": 377425548.0, "step": 9891 }, { "epoch": 1.258364075817326, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.011201858520508, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8662570714950562, "num_tokens": 377462298.0, "step": 9892 }, { "epoch": 1.2584912860959165, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.969534873962402, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8594801425933838, "num_tokens": 377503873.0, "step": 9893 }, { "epoch": 1.258618496374507, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.980121612548828, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8830914497375488, "num_tokens": 377540497.0, "step": 9894 }, { "epoch": 1.2587457066530976, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.0120849609375, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8679602146148682, "num_tokens": 377582242.0, "step": 9895 }, { "epoch": 1.2588729169316881, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.985529899597168, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.883311927318573, "num_tokens": 377614972.0, "step": 9896 }, { "epoch": 1.2590001272102787, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.044013023376465, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8649362921714783, "num_tokens": 377645392.0, "step": 9897 }, { "epoch": 1.2591273374888692, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.977198600769043, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8763425946235657, "num_tokens": 377680317.0, "step": 9898 }, { "epoch": 1.2592545477674597, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.088810920715332, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8675065636634827, "num_tokens": 377714879.0, "step": 9899 }, { "epoch": 1.2593817580460502, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.019420623779297, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.853211522102356, "num_tokens": 377751404.0, "step": 9900 }, { "epoch": 1.2595089683246408, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 12.013761520385742, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.871369481086731, "num_tokens": 377788354.0, "step": 9901 }, { "epoch": 1.259636178603231, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.999273300170898, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8569815158843994, "num_tokens": 377827259.0, "step": 9902 }, { "epoch": 1.2597633888818216, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.029685020446777, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8594231605529785, "num_tokens": 377872477.0, "step": 9903 }, { "epoch": 1.2598905991604121, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.043197631835938, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.860231876373291, "num_tokens": 377911159.0, "step": 9904 }, { "epoch": 1.2600178094390027, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.990614891052246, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8718847036361694, "num_tokens": 377951919.0, "step": 9905 }, { "epoch": 1.2601450197175932, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.005317687988281, "learning_rate": 1e-06, "loss": 0.4737, "mean_token_accuracy": 0.849454402923584, "num_tokens": 377997567.0, "step": 9906 }, { "epoch": 1.2602722299961837, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.959797859191895, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8662456274032593, "num_tokens": 378033811.0, "step": 9907 }, { "epoch": 1.2603994402747742, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.052338600158691, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8549400568008423, "num_tokens": 378070136.0, "step": 9908 }, { "epoch": 1.2605266505533648, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.015700340270996, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8654521703720093, "num_tokens": 378106902.0, "step": 9909 }, { "epoch": 1.2606538608319553, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.992277145385742, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.871708869934082, "num_tokens": 378144638.0, "step": 9910 }, { "epoch": 1.2607810711105456, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.030515670776367, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8765292763710022, "num_tokens": 378179059.0, "step": 9911 }, { "epoch": 1.2609082813891361, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.997343063354492, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8553825616836548, "num_tokens": 378210346.0, "step": 9912 }, { "epoch": 1.2610354916677267, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.029890060424805, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8731908798217773, "num_tokens": 378242687.0, "step": 9913 }, { "epoch": 1.2611627019463172, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.99621868133545, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8737534284591675, "num_tokens": 378279421.0, "step": 9914 }, { "epoch": 1.2612899122249077, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.01233196258545, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8627239465713501, "num_tokens": 378318133.0, "step": 9915 }, { "epoch": 1.2614171225034982, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.043534278869629, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8498886823654175, "num_tokens": 378362987.0, "step": 9916 }, { "epoch": 1.2615443327820888, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.039453506469727, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8844996094703674, "num_tokens": 378396197.0, "step": 9917 }, { "epoch": 1.2616715430606793, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.965487480163574, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8755936622619629, "num_tokens": 378433691.0, "step": 9918 }, { "epoch": 1.2617987533392698, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.050130844116211, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8718981742858887, "num_tokens": 378470277.0, "step": 9919 }, { "epoch": 1.2619259636178604, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.029730796813965, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8615504503250122, "num_tokens": 378510500.0, "step": 9920 }, { "epoch": 1.2620531738964509, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.049026489257812, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8656494617462158, "num_tokens": 378552893.0, "step": 9921 }, { "epoch": 1.2621803841750414, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.045238494873047, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8738267421722412, "num_tokens": 378590917.0, "step": 9922 }, { "epoch": 1.262307594453632, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.950550079345703, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8758954405784607, "num_tokens": 378637764.0, "step": 9923 }, { "epoch": 1.2624348047322225, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.049633026123047, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8638595342636108, "num_tokens": 378675497.0, "step": 9924 }, { "epoch": 1.262562015010813, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.026131629943848, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8588597774505615, "num_tokens": 378710470.0, "step": 9925 }, { "epoch": 1.2626892252894033, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.973191261291504, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8637778162956238, "num_tokens": 378758116.0, "step": 9926 }, { "epoch": 1.2628164355679938, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.015608787536621, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8644605875015259, "num_tokens": 378798166.0, "step": 9927 }, { "epoch": 1.2629436458465844, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.942258834838867, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8825650215148926, "num_tokens": 378839899.0, "step": 9928 }, { "epoch": 1.263070856125175, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.005325317382812, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8679478168487549, "num_tokens": 378886022.0, "step": 9929 }, { "epoch": 1.2631980664037654, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.033119201660156, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8575294613838196, "num_tokens": 378925097.0, "step": 9930 }, { "epoch": 1.263325276682356, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.97183895111084, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8793710470199585, "num_tokens": 378965242.0, "step": 9931 }, { "epoch": 1.2634524869609465, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.0165433883667, "learning_rate": 1e-06, "loss": 0.4857, "mean_token_accuracy": 0.8467649221420288, "num_tokens": 379005932.0, "step": 9932 }, { "epoch": 1.263579697239537, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.032155990600586, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.864501953125, "num_tokens": 379045817.0, "step": 9933 }, { "epoch": 1.2637069075181275, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.00353717803955, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8604490756988525, "num_tokens": 379086481.0, "step": 9934 }, { "epoch": 1.2638341177967178, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.99150276184082, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8501431941986084, "num_tokens": 379129782.0, "step": 9935 }, { "epoch": 1.2639613280753084, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.031989097595215, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8659274578094482, "num_tokens": 379171525.0, "step": 9936 }, { "epoch": 1.264088538353899, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.488432884216309, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8615918159484863, "num_tokens": 379213887.0, "step": 9937 }, { "epoch": 1.2642157486324894, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.91533088684082, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8669941425323486, "num_tokens": 379252634.0, "step": 9938 }, { "epoch": 1.26434295891108, "ewc_loss": 0.026611328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 11.914698600769043, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8662461042404175, "num_tokens": 379287402.0, "step": 9939 }, { "epoch": 1.2644701691896705, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.975082397460938, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8725199699401855, "num_tokens": 379325592.0, "step": 9940 }, { "epoch": 1.264597379468261, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.944491386413574, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8552138805389404, "num_tokens": 379359010.0, "step": 9941 }, { "epoch": 1.2647245897468515, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.979507446289062, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8595377206802368, "num_tokens": 379399783.0, "step": 9942 }, { "epoch": 1.264851800025442, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.928166389465332, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8703644871711731, "num_tokens": 379442207.0, "step": 9943 }, { "epoch": 1.2649790103040326, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.012001991271973, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8668779134750366, "num_tokens": 379479561.0, "step": 9944 }, { "epoch": 1.2651062205826231, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.995521545410156, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8723974823951721, "num_tokens": 379516413.0, "step": 9945 }, { "epoch": 1.2652334308612136, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.978018760681152, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8582119941711426, "num_tokens": 379557993.0, "step": 9946 }, { "epoch": 1.2653606411398042, "ewc_loss": 0.02685546875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.962015151977539, "learning_rate": 1e-06, "loss": 0.4726, "mean_token_accuracy": 0.8491727113723755, "num_tokens": 379594028.0, "step": 9947 }, { "epoch": 1.2654878514183947, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.990300178527832, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8613823652267456, "num_tokens": 379631235.0, "step": 9948 }, { "epoch": 1.2656150616969852, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.990571022033691, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8554335832595825, "num_tokens": 379668714.0, "step": 9949 }, { "epoch": 1.2657422719755758, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.975544929504395, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8721165060997009, "num_tokens": 379710233.0, "step": 9950 }, { "epoch": 1.265869482254166, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.966586112976074, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8642705678939819, "num_tokens": 379746208.0, "step": 9951 }, { "epoch": 1.2659966925327566, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.975204467773438, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8761480450630188, "num_tokens": 379796420.0, "step": 9952 }, { "epoch": 1.2661239028113471, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.983019828796387, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8679603338241577, "num_tokens": 379830027.0, "step": 9953 }, { "epoch": 1.2662511130899377, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.990311622619629, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8669039607048035, "num_tokens": 379871751.0, "step": 9954 }, { "epoch": 1.2663783233685282, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.958540916442871, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8601061105728149, "num_tokens": 379908255.0, "step": 9955 }, { "epoch": 1.2665055336471187, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.004796981811523, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8684206008911133, "num_tokens": 379944502.0, "step": 9956 }, { "epoch": 1.2666327439257092, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.913812637329102, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8637576103210449, "num_tokens": 379981929.0, "step": 9957 }, { "epoch": 1.2667599542042998, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.988428115844727, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8619177937507629, "num_tokens": 380018876.0, "step": 9958 }, { "epoch": 1.2668871644828903, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.971397399902344, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.86187744140625, "num_tokens": 380054153.0, "step": 9959 }, { "epoch": 1.2670143747614806, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.020188331604004, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.858542799949646, "num_tokens": 380086550.0, "step": 9960 }, { "epoch": 1.2671415850400711, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.9585542678833, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8615609407424927, "num_tokens": 380120308.0, "step": 9961 }, { "epoch": 1.2672687953186617, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.963859558105469, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8502035140991211, "num_tokens": 380160204.0, "step": 9962 }, { "epoch": 1.2673960055972522, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.004719734191895, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8700584173202515, "num_tokens": 380196687.0, "step": 9963 }, { "epoch": 1.2675232158758427, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.989048957824707, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8813756704330444, "num_tokens": 380237451.0, "step": 9964 }, { "epoch": 1.2676504261544332, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.973316192626953, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8626585006713867, "num_tokens": 380276631.0, "step": 9965 }, { "epoch": 1.2677776364330238, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.964442253112793, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.871904194355011, "num_tokens": 380319432.0, "step": 9966 }, { "epoch": 1.2679048467116143, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.959478378295898, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8741881847381592, "num_tokens": 380356666.0, "step": 9967 }, { "epoch": 1.2680320569902048, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.035008430480957, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.874923825263977, "num_tokens": 380390347.0, "step": 9968 }, { "epoch": 1.2681592672687954, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.010310173034668, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8701127767562866, "num_tokens": 380424083.0, "step": 9969 }, { "epoch": 1.2682864775473859, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.997337341308594, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.860690712928772, "num_tokens": 380471179.0, "step": 9970 }, { "epoch": 1.2684136878259764, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.998892784118652, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8675772547721863, "num_tokens": 380510291.0, "step": 9971 }, { "epoch": 1.268540898104567, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.958358764648438, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8715358972549438, "num_tokens": 380556297.0, "step": 9972 }, { "epoch": 1.2686681083831575, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.031262397766113, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8770208358764648, "num_tokens": 380592113.0, "step": 9973 }, { "epoch": 1.268795318661748, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.032083511352539, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8791718482971191, "num_tokens": 380623942.0, "step": 9974 }, { "epoch": 1.2689225289403383, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.973143577575684, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8605090379714966, "num_tokens": 380662300.0, "step": 9975 }, { "epoch": 1.2690497392189288, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.007943153381348, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.868689775466919, "num_tokens": 380701452.0, "step": 9976 }, { "epoch": 1.2691769494975194, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.008628845214844, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8775268197059631, "num_tokens": 380741248.0, "step": 9977 }, { "epoch": 1.2693041597761099, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.021604537963867, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8603477478027344, "num_tokens": 380781455.0, "step": 9978 }, { "epoch": 1.2694313700547004, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.987509727478027, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8590307235717773, "num_tokens": 380818139.0, "step": 9979 }, { "epoch": 1.269558580333291, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.02224349975586, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8718308210372925, "num_tokens": 380860530.0, "step": 9980 }, { "epoch": 1.2696857906118815, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.949376106262207, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.878737211227417, "num_tokens": 380900825.0, "step": 9981 }, { "epoch": 1.269813000890472, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.984650611877441, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8641033172607422, "num_tokens": 380939270.0, "step": 9982 }, { "epoch": 1.2699402111690625, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.99312686920166, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.844197154045105, "num_tokens": 380981708.0, "step": 9983 }, { "epoch": 1.2700674214476528, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.008988380432129, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8722608685493469, "num_tokens": 381015186.0, "step": 9984 }, { "epoch": 1.2701946317262434, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.951089859008789, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8767271041870117, "num_tokens": 381055503.0, "step": 9985 }, { "epoch": 1.270321842004834, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.997184753417969, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8797726631164551, "num_tokens": 381091935.0, "step": 9986 }, { "epoch": 1.2704490522834244, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.043027877807617, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8717596530914307, "num_tokens": 381127951.0, "step": 9987 }, { "epoch": 1.270576262562015, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.000539779663086, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8618460893630981, "num_tokens": 381161918.0, "step": 9988 }, { "epoch": 1.2707034728406055, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.021240234375, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8558799028396606, "num_tokens": 381198747.0, "step": 9989 }, { "epoch": 1.270830683119196, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.952326774597168, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8684590458869934, "num_tokens": 381242714.0, "step": 9990 }, { "epoch": 1.2709578933977865, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.025146484375, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8794490098953247, "num_tokens": 381274667.0, "step": 9991 }, { "epoch": 1.271085103676377, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.981003761291504, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.877112090587616, "num_tokens": 381318110.0, "step": 9992 }, { "epoch": 1.2712123139549676, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.960678100585938, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8684155941009521, "num_tokens": 381356694.0, "step": 9993 }, { "epoch": 1.2713395242335581, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.014183044433594, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8568017482757568, "num_tokens": 381395218.0, "step": 9994 }, { "epoch": 1.2714667345121486, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.994528770446777, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8803911209106445, "num_tokens": 381429195.0, "step": 9995 }, { "epoch": 1.2715939447907392, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.007436752319336, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8614057302474976, "num_tokens": 381468276.0, "step": 9996 }, { "epoch": 1.2717211550693297, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.021846771240234, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.875578761100769, "num_tokens": 381503400.0, "step": 9997 }, { "epoch": 1.2718483653479202, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.994449615478516, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8729643821716309, "num_tokens": 381542126.0, "step": 9998 }, { "epoch": 1.2719755756265108, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.985915184020996, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8759950399398804, "num_tokens": 381579169.0, "step": 9999 }, { "epoch": 1.272102785905101, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.033905982971191, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8525367379188538, "num_tokens": 381615501.0, "step": 10000 }, { "epoch": 1.2722299961836916, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.0165376663208, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8618675470352173, "num_tokens": 381649737.0, "step": 10001 }, { "epoch": 1.2723572064622821, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.007326126098633, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8704614639282227, "num_tokens": 381693776.0, "step": 10002 }, { "epoch": 1.2724844167408726, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.019139289855957, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8772106766700745, "num_tokens": 381733324.0, "step": 10003 }, { "epoch": 1.2726116270194632, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.019574165344238, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8635983467102051, "num_tokens": 381765893.0, "step": 10004 }, { "epoch": 1.2727388372980537, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.945989608764648, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8653221130371094, "num_tokens": 381802782.0, "step": 10005 }, { "epoch": 1.2728660475766442, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.006904602050781, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8703106641769409, "num_tokens": 381843448.0, "step": 10006 }, { "epoch": 1.2729932578552348, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.991896629333496, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.856250524520874, "num_tokens": 381878850.0, "step": 10007 }, { "epoch": 1.2731204681338253, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.981376647949219, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.8468872308731079, "num_tokens": 381915508.0, "step": 10008 }, { "epoch": 1.2732476784124156, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.996747970581055, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8575426340103149, "num_tokens": 381952699.0, "step": 10009 }, { "epoch": 1.2733748886910061, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.014923095703125, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8601784706115723, "num_tokens": 381989947.0, "step": 10010 }, { "epoch": 1.2735020989695967, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.023737907409668, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8665802478790283, "num_tokens": 382027701.0, "step": 10011 }, { "epoch": 1.2736293092481872, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.05052661895752, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8659611940383911, "num_tokens": 382066415.0, "step": 10012 }, { "epoch": 1.2737565195267777, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.988255500793457, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8622259497642517, "num_tokens": 382108402.0, "step": 10013 }, { "epoch": 1.2738837298053682, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.063826560974121, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8673697113990784, "num_tokens": 382146865.0, "step": 10014 }, { "epoch": 1.2740109400839588, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.021198272705078, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8642565011978149, "num_tokens": 382186746.0, "step": 10015 }, { "epoch": 1.2741381503625493, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.014328002929688, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.880799412727356, "num_tokens": 382220693.0, "step": 10016 }, { "epoch": 1.2742653606411398, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.003081321716309, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8604768514633179, "num_tokens": 382259549.0, "step": 10017 }, { "epoch": 1.2743925709197303, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.993475914001465, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8768811821937561, "num_tokens": 382299164.0, "step": 10018 }, { "epoch": 1.2745197811983209, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.065447807312012, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8799289464950562, "num_tokens": 382333596.0, "step": 10019 }, { "epoch": 1.2746469914769114, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.063836097717285, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8690099716186523, "num_tokens": 382372321.0, "step": 10020 }, { "epoch": 1.274774201755502, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.009382247924805, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8695332407951355, "num_tokens": 382411125.0, "step": 10021 }, { "epoch": 1.2749014120340925, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.050165176391602, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.857830286026001, "num_tokens": 382450935.0, "step": 10022 }, { "epoch": 1.275028622312683, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.022966384887695, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8649217486381531, "num_tokens": 382488232.0, "step": 10023 }, { "epoch": 1.2751558325912733, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.04641056060791, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8620965480804443, "num_tokens": 382522280.0, "step": 10024 }, { "epoch": 1.2752830428698638, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.01296329498291, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8775684833526611, "num_tokens": 382562788.0, "step": 10025 }, { "epoch": 1.2754102531484544, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.995696067810059, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8598973751068115, "num_tokens": 382598534.0, "step": 10026 }, { "epoch": 1.2755374634270449, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.07454776763916, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8612281084060669, "num_tokens": 382641803.0, "step": 10027 }, { "epoch": 1.2756646737056354, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.962223052978516, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8779354095458984, "num_tokens": 382677871.0, "step": 10028 }, { "epoch": 1.275791883984226, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.074213027954102, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8765551447868347, "num_tokens": 382714989.0, "step": 10029 }, { "epoch": 1.2759190942628165, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.950965881347656, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8706380128860474, "num_tokens": 382751919.0, "step": 10030 }, { "epoch": 1.276046304541407, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.054584503173828, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8785515427589417, "num_tokens": 382793550.0, "step": 10031 }, { "epoch": 1.2761735148199975, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.968551635742188, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8747411966323853, "num_tokens": 382833142.0, "step": 10032 }, { "epoch": 1.2763007250985878, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.046972274780273, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8749812841415405, "num_tokens": 382869934.0, "step": 10033 }, { "epoch": 1.2764279353771784, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.000554084777832, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8539751768112183, "num_tokens": 382909533.0, "step": 10034 }, { "epoch": 1.2765551456557689, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.028428077697754, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8730214834213257, "num_tokens": 382947462.0, "step": 10035 }, { "epoch": 1.2766823559343594, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.997227668762207, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8690296411514282, "num_tokens": 382989973.0, "step": 10036 }, { "epoch": 1.27680956621295, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.991790771484375, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8656424283981323, "num_tokens": 383033052.0, "step": 10037 }, { "epoch": 1.2769367764915405, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.99938678741455, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8660132884979248, "num_tokens": 383075981.0, "step": 10038 }, { "epoch": 1.277063986770131, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.998461723327637, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8840984106063843, "num_tokens": 383111984.0, "step": 10039 }, { "epoch": 1.2771911970487215, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.01474380493164, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8681226968765259, "num_tokens": 383155323.0, "step": 10040 }, { "epoch": 1.277318407327312, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.004911422729492, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.86664879322052, "num_tokens": 383190069.0, "step": 10041 }, { "epoch": 1.2774456176059026, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.007294654846191, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8586733937263489, "num_tokens": 383224999.0, "step": 10042 }, { "epoch": 1.2775728278844931, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.063207626342773, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8598687648773193, "num_tokens": 383267764.0, "step": 10043 }, { "epoch": 1.2777000381630836, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.971488952636719, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8764581680297852, "num_tokens": 383306148.0, "step": 10044 }, { "epoch": 1.2778272484416742, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.983002662658691, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8714004755020142, "num_tokens": 383343408.0, "step": 10045 }, { "epoch": 1.2779544587202647, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.945109367370605, "learning_rate": 1e-06, "loss": 0.3344, "mean_token_accuracy": 0.8906149864196777, "num_tokens": 383383626.0, "step": 10046 }, { "epoch": 1.2780816689988552, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.976041793823242, "learning_rate": 1e-06, "loss": 0.4751, "mean_token_accuracy": 0.8518158197402954, "num_tokens": 383423218.0, "step": 10047 }, { "epoch": 1.2782088792774458, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.992630004882812, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8745291233062744, "num_tokens": 383461126.0, "step": 10048 }, { "epoch": 1.278336089556036, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.952634811401367, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8590072989463806, "num_tokens": 383501808.0, "step": 10049 }, { "epoch": 1.2784632998346266, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.010064125061035, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8879733681678772, "num_tokens": 383542395.0, "step": 10050 }, { "epoch": 1.2785905101132171, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.971978187561035, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8641072511672974, "num_tokens": 383579968.0, "step": 10051 }, { "epoch": 1.2787177203918076, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.013675689697266, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8664953708648682, "num_tokens": 383617861.0, "step": 10052 }, { "epoch": 1.2788449306703982, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.980915069580078, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8628942966461182, "num_tokens": 383652534.0, "step": 10053 }, { "epoch": 1.2789721409489887, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.986468315124512, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8626112341880798, "num_tokens": 383693120.0, "step": 10054 }, { "epoch": 1.2790993512275792, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.060774803161621, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8579241037368774, "num_tokens": 383734958.0, "step": 10055 }, { "epoch": 1.2792265615061698, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 11.999796867370605, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8604936599731445, "num_tokens": 383768873.0, "step": 10056 }, { "epoch": 1.2793537717847603, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.10327434539795, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8687918186187744, "num_tokens": 383804063.0, "step": 10057 }, { "epoch": 1.2794809820633506, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.959113121032715, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8812751173973083, "num_tokens": 383843098.0, "step": 10058 }, { "epoch": 1.2796081923419411, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.070204734802246, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.875960111618042, "num_tokens": 383879443.0, "step": 10059 }, { "epoch": 1.2797354026205316, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.991451263427734, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8666504621505737, "num_tokens": 383918071.0, "step": 10060 }, { "epoch": 1.2798626128991222, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 12.003766059875488, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8610248565673828, "num_tokens": 383953052.0, "step": 10061 }, { "epoch": 1.2799898231777127, "ewc_loss": 0.027099609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.013327598571777, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.865007758140564, "num_tokens": 383992835.0, "step": 10062 }, { "epoch": 1.2801170334563032, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.042437553405762, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8545578718185425, "num_tokens": 384031234.0, "step": 10063 }, { "epoch": 1.2802442437348938, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.059266090393066, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8589255809783936, "num_tokens": 384063611.0, "step": 10064 }, { "epoch": 1.2803714540134843, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.009098052978516, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8752985596656799, "num_tokens": 384106757.0, "step": 10065 }, { "epoch": 1.2804986642920748, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.000336647033691, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.871665358543396, "num_tokens": 384143094.0, "step": 10066 }, { "epoch": 1.2806258745706653, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.993797302246094, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8677574396133423, "num_tokens": 384181498.0, "step": 10067 }, { "epoch": 1.2807530848492559, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.025614738464355, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8719840049743652, "num_tokens": 384209784.0, "step": 10068 }, { "epoch": 1.2808802951278464, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.983792304992676, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8732140064239502, "num_tokens": 384248000.0, "step": 10069 }, { "epoch": 1.281007505406437, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.037294387817383, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.857978343963623, "num_tokens": 384290897.0, "step": 10070 }, { "epoch": 1.2811347156850275, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.02005386352539, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8610180020332336, "num_tokens": 384329464.0, "step": 10071 }, { "epoch": 1.281261925963618, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.011055946350098, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8708198070526123, "num_tokens": 384368915.0, "step": 10072 }, { "epoch": 1.2813891362422083, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.045462608337402, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8771493434906006, "num_tokens": 384407445.0, "step": 10073 }, { "epoch": 1.2815163465207988, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.000102043151855, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8637810945510864, "num_tokens": 384444384.0, "step": 10074 }, { "epoch": 1.2816435567993893, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.03682804107666, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8642001152038574, "num_tokens": 384479670.0, "step": 10075 }, { "epoch": 1.2817707670779799, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.036697387695312, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8560458421707153, "num_tokens": 384522436.0, "step": 10076 }, { "epoch": 1.2818979773565704, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.058560371398926, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.860489547252655, "num_tokens": 384559754.0, "step": 10077 }, { "epoch": 1.282025187635161, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.024675369262695, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.864748477935791, "num_tokens": 384595153.0, "step": 10078 }, { "epoch": 1.2821523979137515, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.061454772949219, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8667402267456055, "num_tokens": 384628345.0, "step": 10079 }, { "epoch": 1.282279608192342, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.071014404296875, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8626477718353271, "num_tokens": 384666725.0, "step": 10080 }, { "epoch": 1.2824068184709325, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.026957511901855, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8581933975219727, "num_tokens": 384706217.0, "step": 10081 }, { "epoch": 1.2825340287495228, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.027327537536621, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8640315532684326, "num_tokens": 384741835.0, "step": 10082 }, { "epoch": 1.2826612390281134, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.01099681854248, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8679032325744629, "num_tokens": 384777273.0, "step": 10083 }, { "epoch": 1.2827884493067039, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.982376098632812, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8811137676239014, "num_tokens": 384814950.0, "step": 10084 }, { "epoch": 1.2829156595852944, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.094627380371094, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8594595193862915, "num_tokens": 384848088.0, "step": 10085 }, { "epoch": 1.283042869863885, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.010638236999512, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8606232404708862, "num_tokens": 384887361.0, "step": 10086 }, { "epoch": 1.2831700801424755, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.989716529846191, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8733668923377991, "num_tokens": 384930459.0, "step": 10087 }, { "epoch": 1.283297290421066, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.063694953918457, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8611323833465576, "num_tokens": 384969506.0, "step": 10088 }, { "epoch": 1.2834245006996565, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.008886337280273, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8734889030456543, "num_tokens": 385004368.0, "step": 10089 }, { "epoch": 1.283551710978247, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.066767692565918, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8689465522766113, "num_tokens": 385044553.0, "step": 10090 }, { "epoch": 1.2836789212568376, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.999526023864746, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8789905309677124, "num_tokens": 385076873.0, "step": 10091 }, { "epoch": 1.283806131535428, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.00107192993164, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8691359758377075, "num_tokens": 385113969.0, "step": 10092 }, { "epoch": 1.2839333418140186, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.994898796081543, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.877680778503418, "num_tokens": 385149460.0, "step": 10093 }, { "epoch": 1.2840605520926092, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.095233917236328, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8504235148429871, "num_tokens": 385187240.0, "step": 10094 }, { "epoch": 1.2841877623711997, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.07059097290039, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8592104911804199, "num_tokens": 385225532.0, "step": 10095 }, { "epoch": 1.2843149726497902, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.025382041931152, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8699156641960144, "num_tokens": 385267404.0, "step": 10096 }, { "epoch": 1.2844421829283807, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.089518547058105, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.855812668800354, "num_tokens": 385314008.0, "step": 10097 }, { "epoch": 1.284569393206971, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.034464836120605, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8550761938095093, "num_tokens": 385352204.0, "step": 10098 }, { "epoch": 1.2846966034855616, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.069601058959961, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8593224883079529, "num_tokens": 385393307.0, "step": 10099 }, { "epoch": 1.2848238137641521, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.079487800598145, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8685071468353271, "num_tokens": 385426124.0, "step": 10100 }, { "epoch": 1.2849510240427426, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.992327690124512, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8723558187484741, "num_tokens": 385467471.0, "step": 10101 }, { "epoch": 1.2850782343213332, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.128190040588379, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8572713136672974, "num_tokens": 385508780.0, "step": 10102 }, { "epoch": 1.2852054445999237, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.034205436706543, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8722359538078308, "num_tokens": 385552646.0, "step": 10103 }, { "epoch": 1.2853326548785142, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.067244529724121, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8549971580505371, "num_tokens": 385586676.0, "step": 10104 }, { "epoch": 1.2854598651571048, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.201094627380371, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8555557131767273, "num_tokens": 385623201.0, "step": 10105 }, { "epoch": 1.2855870754356953, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 11.965014457702637, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8779221773147583, "num_tokens": 385662716.0, "step": 10106 }, { "epoch": 1.2857142857142856, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.038193702697754, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8569080829620361, "num_tokens": 385706986.0, "step": 10107 }, { "epoch": 1.2858414959928761, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.12035083770752, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.866277277469635, "num_tokens": 385744240.0, "step": 10108 }, { "epoch": 1.2859687062714666, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.055159568786621, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8512143492698669, "num_tokens": 385778583.0, "step": 10109 }, { "epoch": 1.2860959165500572, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.999106407165527, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8678203821182251, "num_tokens": 385819910.0, "step": 10110 }, { "epoch": 1.2862231268286477, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.01283073425293, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8645057678222656, "num_tokens": 385860068.0, "step": 10111 }, { "epoch": 1.2863503371072382, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.05989933013916, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8588087558746338, "num_tokens": 385890460.0, "step": 10112 }, { "epoch": 1.2864775473858288, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.074408531188965, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8638286590576172, "num_tokens": 385926174.0, "step": 10113 }, { "epoch": 1.2866047576644193, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.04012393951416, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8765504360198975, "num_tokens": 385962928.0, "step": 10114 }, { "epoch": 1.2867319679430098, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.034400939941406, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8747955560684204, "num_tokens": 386002509.0, "step": 10115 }, { "epoch": 1.2868591782216003, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.993149757385254, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8660516738891602, "num_tokens": 386043304.0, "step": 10116 }, { "epoch": 1.2869863885001909, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.08324909210205, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8736187219619751, "num_tokens": 386081607.0, "step": 10117 }, { "epoch": 1.2871135987787814, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.001837730407715, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8862545490264893, "num_tokens": 386117762.0, "step": 10118 }, { "epoch": 1.287240809057372, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.96946907043457, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8592721819877625, "num_tokens": 386155908.0, "step": 10119 }, { "epoch": 1.2873680193359625, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.047131538391113, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8650039434432983, "num_tokens": 386199385.0, "step": 10120 }, { "epoch": 1.287495229614553, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.0502347946167, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8621249198913574, "num_tokens": 386231892.0, "step": 10121 }, { "epoch": 1.2876224398931433, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.020953178405762, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8510445952415466, "num_tokens": 386269657.0, "step": 10122 }, { "epoch": 1.2877496501717338, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.97395133972168, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8722363710403442, "num_tokens": 386306936.0, "step": 10123 }, { "epoch": 1.2878768604503243, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.04403018951416, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8677628040313721, "num_tokens": 386342260.0, "step": 10124 }, { "epoch": 1.2880040707289149, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.02701187133789, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8623111248016357, "num_tokens": 386381784.0, "step": 10125 }, { "epoch": 1.2881312810075054, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.041553497314453, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8656463623046875, "num_tokens": 386422350.0, "step": 10126 }, { "epoch": 1.288258491286096, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.071971893310547, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8626531362533569, "num_tokens": 386467951.0, "step": 10127 }, { "epoch": 1.2883857015646865, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.997806549072266, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8760370016098022, "num_tokens": 386511896.0, "step": 10128 }, { "epoch": 1.288512911843277, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.034931182861328, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8542959094047546, "num_tokens": 386550704.0, "step": 10129 }, { "epoch": 1.2886401221218675, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.032883644104004, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8502102494239807, "num_tokens": 386589295.0, "step": 10130 }, { "epoch": 1.2887673324004578, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.086918830871582, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8664531111717224, "num_tokens": 386623462.0, "step": 10131 }, { "epoch": 1.2888945426790483, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.998488426208496, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8697252869606018, "num_tokens": 386663036.0, "step": 10132 }, { "epoch": 1.2890217529576389, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.96998119354248, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8666219711303711, "num_tokens": 386702530.0, "step": 10133 }, { "epoch": 1.2891489632362294, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.027599334716797, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8814094066619873, "num_tokens": 386739758.0, "step": 10134 }, { "epoch": 1.28927617351482, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.978829383850098, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8767940998077393, "num_tokens": 386778287.0, "step": 10135 }, { "epoch": 1.2894033837934105, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.053686141967773, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8773478269577026, "num_tokens": 386812348.0, "step": 10136 }, { "epoch": 1.289530594072001, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.998559951782227, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8741733431816101, "num_tokens": 386855540.0, "step": 10137 }, { "epoch": 1.2896578043505915, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.037803649902344, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8615849614143372, "num_tokens": 386893775.0, "step": 10138 }, { "epoch": 1.289785014629182, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.04680347442627, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8672366142272949, "num_tokens": 386929553.0, "step": 10139 }, { "epoch": 1.2899122249077726, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.99982738494873, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8614374399185181, "num_tokens": 386968033.0, "step": 10140 }, { "epoch": 1.290039435186363, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.04784107208252, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.865685224533081, "num_tokens": 387004967.0, "step": 10141 }, { "epoch": 1.2901666454649536, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.038826942443848, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8550922870635986, "num_tokens": 387038086.0, "step": 10142 }, { "epoch": 1.2902938557435442, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.996582984924316, "learning_rate": 1e-06, "loss": 0.499, "mean_token_accuracy": 0.8415676355361938, "num_tokens": 387077112.0, "step": 10143 }, { "epoch": 1.2904210660221347, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.019912719726562, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8505412340164185, "num_tokens": 387118800.0, "step": 10144 }, { "epoch": 1.2905482763007252, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.043183326721191, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8615944385528564, "num_tokens": 387160321.0, "step": 10145 }, { "epoch": 1.2906754865793157, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.001797676086426, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8644627332687378, "num_tokens": 387199387.0, "step": 10146 }, { "epoch": 1.290802696857906, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.032505989074707, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8721315860748291, "num_tokens": 387238859.0, "step": 10147 }, { "epoch": 1.2909299071364966, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 11.99856185913086, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.87183678150177, "num_tokens": 387272149.0, "step": 10148 }, { "epoch": 1.291057117415087, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.05285358428955, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.864967942237854, "num_tokens": 387311617.0, "step": 10149 }, { "epoch": 1.2911843276936776, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 12.038236618041992, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8532484769821167, "num_tokens": 387344281.0, "step": 10150 }, { "epoch": 1.2913115379722682, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.0270357131958, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8559836149215698, "num_tokens": 387385959.0, "step": 10151 }, { "epoch": 1.2914387482508587, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.030580520629883, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8612688779830933, "num_tokens": 387426197.0, "step": 10152 }, { "epoch": 1.2915659585294492, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.050551414489746, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8718374967575073, "num_tokens": 387462250.0, "step": 10153 }, { "epoch": 1.2916931688080397, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.030545234680176, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.866203248500824, "num_tokens": 387499903.0, "step": 10154 }, { "epoch": 1.2918203790866303, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.028560638427734, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8728825449943542, "num_tokens": 387535361.0, "step": 10155 }, { "epoch": 1.2919475893652206, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.021254539489746, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8629817962646484, "num_tokens": 387574300.0, "step": 10156 }, { "epoch": 1.2920747996438111, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.062829971313477, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.868285059928894, "num_tokens": 387606860.0, "step": 10157 }, { "epoch": 1.2922020099224016, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.04354190826416, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8570855855941772, "num_tokens": 387646519.0, "step": 10158 }, { "epoch": 1.2923292202009922, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.08883285522461, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8780324459075928, "num_tokens": 387681100.0, "step": 10159 }, { "epoch": 1.2924564304795827, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.017071723937988, "learning_rate": 1e-06, "loss": 0.3537, "mean_token_accuracy": 0.8881436586380005, "num_tokens": 387715962.0, "step": 10160 }, { "epoch": 1.2925836407581732, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.057075500488281, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8645645380020142, "num_tokens": 387755856.0, "step": 10161 }, { "epoch": 1.2927108510367638, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.06724739074707, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8746495246887207, "num_tokens": 387793256.0, "step": 10162 }, { "epoch": 1.2928380613153543, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.028510093688965, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8628994226455688, "num_tokens": 387834696.0, "step": 10163 }, { "epoch": 1.2929652715939448, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.082818031311035, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.85537189245224, "num_tokens": 387874973.0, "step": 10164 }, { "epoch": 1.2930924818725353, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.117424011230469, "learning_rate": 1e-06, "loss": 0.4579, "mean_token_accuracy": 0.8566410541534424, "num_tokens": 387904887.0, "step": 10165 }, { "epoch": 1.2932196921511259, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.043943405151367, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8738059997558594, "num_tokens": 387944676.0, "step": 10166 }, { "epoch": 1.2933469024297164, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.062628746032715, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8675091862678528, "num_tokens": 387982234.0, "step": 10167 }, { "epoch": 1.293474112708307, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.09558391571045, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8685938715934753, "num_tokens": 388018001.0, "step": 10168 }, { "epoch": 1.2936013229868975, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.052509307861328, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8618555068969727, "num_tokens": 388057476.0, "step": 10169 }, { "epoch": 1.293728533265488, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.108222007751465, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8845884799957275, "num_tokens": 388095351.0, "step": 10170 }, { "epoch": 1.2938557435440783, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.044296264648438, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8659422397613525, "num_tokens": 388135366.0, "step": 10171 }, { "epoch": 1.2939829538226688, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.081969261169434, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8558206558227539, "num_tokens": 388175003.0, "step": 10172 }, { "epoch": 1.2941101641012593, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.104642868041992, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8483173847198486, "num_tokens": 388207083.0, "step": 10173 }, { "epoch": 1.2942373743798499, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.035747528076172, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8824946880340576, "num_tokens": 388245038.0, "step": 10174 }, { "epoch": 1.2943645846584404, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.036155700683594, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8626095056533813, "num_tokens": 388285035.0, "step": 10175 }, { "epoch": 1.294491794937031, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.006834030151367, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8719191551208496, "num_tokens": 388323971.0, "step": 10176 }, { "epoch": 1.2946190052156215, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.110784530639648, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8620781898498535, "num_tokens": 388362256.0, "step": 10177 }, { "epoch": 1.294746215494212, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.023351669311523, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8861960768699646, "num_tokens": 388401120.0, "step": 10178 }, { "epoch": 1.2948734257728025, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.059470176696777, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8693579435348511, "num_tokens": 388446833.0, "step": 10179 }, { "epoch": 1.2950006360513928, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.084800720214844, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8794103860855103, "num_tokens": 388482266.0, "step": 10180 }, { "epoch": 1.2951278463299833, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.073811531066895, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8680589199066162, "num_tokens": 388518024.0, "step": 10181 }, { "epoch": 1.2952550566085739, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.070680618286133, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8692059516906738, "num_tokens": 388564246.0, "step": 10182 }, { "epoch": 1.2953822668871644, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.015307426452637, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8710494041442871, "num_tokens": 388603556.0, "step": 10183 }, { "epoch": 1.295509477165755, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.089530944824219, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8753895163536072, "num_tokens": 388643615.0, "step": 10184 }, { "epoch": 1.2956366874443455, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.037818908691406, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8704789876937866, "num_tokens": 388682917.0, "step": 10185 }, { "epoch": 1.295763897722936, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.10184383392334, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8697868585586548, "num_tokens": 388720269.0, "step": 10186 }, { "epoch": 1.2958911080015265, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.028034210205078, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8602570295333862, "num_tokens": 388759245.0, "step": 10187 }, { "epoch": 1.296018318280117, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.096525192260742, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8640292882919312, "num_tokens": 388795673.0, "step": 10188 }, { "epoch": 1.2961455285587076, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.019712448120117, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8711308240890503, "num_tokens": 388832926.0, "step": 10189 }, { "epoch": 1.296272738837298, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.029213905334473, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8710371255874634, "num_tokens": 388869767.0, "step": 10190 }, { "epoch": 1.2963999491158886, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.063681602478027, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8697083592414856, "num_tokens": 388913347.0, "step": 10191 }, { "epoch": 1.2965271593944792, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.005017280578613, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8793932199478149, "num_tokens": 388948211.0, "step": 10192 }, { "epoch": 1.2966543696730697, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.079591751098633, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8721557855606079, "num_tokens": 388989198.0, "step": 10193 }, { "epoch": 1.2967815799516602, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.016305923461914, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8585887551307678, "num_tokens": 389029076.0, "step": 10194 }, { "epoch": 1.2969087902302507, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.09423542022705, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8684442043304443, "num_tokens": 389066751.0, "step": 10195 }, { "epoch": 1.297036000508841, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.054612159729004, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8657358884811401, "num_tokens": 389104207.0, "step": 10196 }, { "epoch": 1.2971632107874316, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.005663871765137, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.870930552482605, "num_tokens": 389141941.0, "step": 10197 }, { "epoch": 1.297290421066022, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.0899658203125, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.877307653427124, "num_tokens": 389185326.0, "step": 10198 }, { "epoch": 1.2974176313446126, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.01694393157959, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8769137859344482, "num_tokens": 389221512.0, "step": 10199 }, { "epoch": 1.2975448416232032, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.045378684997559, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8751507997512817, "num_tokens": 389259388.0, "step": 10200 }, { "epoch": 1.2976720519017937, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.998764991760254, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8564748764038086, "num_tokens": 389298773.0, "step": 10201 }, { "epoch": 1.2977992621803842, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.04049301147461, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.868682861328125, "num_tokens": 389339680.0, "step": 10202 }, { "epoch": 1.2979264724589747, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.062098503112793, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8708922863006592, "num_tokens": 389380755.0, "step": 10203 }, { "epoch": 1.2980536827375653, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.061644554138184, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8732110261917114, "num_tokens": 389412031.0, "step": 10204 }, { "epoch": 1.2981808930161556, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.031476974487305, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8774994015693665, "num_tokens": 389448299.0, "step": 10205 }, { "epoch": 1.298308103294746, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.096839904785156, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8622974753379822, "num_tokens": 389483962.0, "step": 10206 }, { "epoch": 1.2984353135733366, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 11.995109558105469, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8602339029312134, "num_tokens": 389521989.0, "step": 10207 }, { "epoch": 1.2985625238519272, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.116368293762207, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8624840974807739, "num_tokens": 389558228.0, "step": 10208 }, { "epoch": 1.2986897341305177, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.02943229675293, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8673245906829834, "num_tokens": 389597164.0, "step": 10209 }, { "epoch": 1.2988169444091082, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.058170318603516, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.867559552192688, "num_tokens": 389636435.0, "step": 10210 }, { "epoch": 1.2989441546876987, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.098578453063965, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8603053092956543, "num_tokens": 389670498.0, "step": 10211 }, { "epoch": 1.2990713649662893, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.091715812683105, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8698306083679199, "num_tokens": 389707433.0, "step": 10212 }, { "epoch": 1.2991985752448798, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105015754699707, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8676841259002686, "num_tokens": 389747300.0, "step": 10213 }, { "epoch": 1.2993257855234703, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.065664291381836, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8622745275497437, "num_tokens": 389788381.0, "step": 10214 }, { "epoch": 1.2994529958020609, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.095978736877441, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8716152310371399, "num_tokens": 389821792.0, "step": 10215 }, { "epoch": 1.2995802060806514, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.020317077636719, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8792340755462646, "num_tokens": 389859370.0, "step": 10216 }, { "epoch": 1.299707416359242, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.085946083068848, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.858657717704773, "num_tokens": 389897020.0, "step": 10217 }, { "epoch": 1.2998346266378324, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.071891784667969, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8415759801864624, "num_tokens": 389927107.0, "step": 10218 }, { "epoch": 1.299961836916423, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.115571022033691, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8647228479385376, "num_tokens": 389966323.0, "step": 10219 }, { "epoch": 1.3000890471950133, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.034570693969727, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8672041296958923, "num_tokens": 390001416.0, "step": 10220 }, { "epoch": 1.3002162574736038, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.083921432495117, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8704306483268738, "num_tokens": 390042326.0, "step": 10221 }, { "epoch": 1.3003434677521943, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.065556526184082, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8576165437698364, "num_tokens": 390084443.0, "step": 10222 }, { "epoch": 1.3004706780307849, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.03133773803711, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8608239889144897, "num_tokens": 390126309.0, "step": 10223 }, { "epoch": 1.3005978883093754, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.04288101196289, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8720308542251587, "num_tokens": 390161463.0, "step": 10224 }, { "epoch": 1.300725098587966, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.02658748626709, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8780096173286438, "num_tokens": 390203221.0, "step": 10225 }, { "epoch": 1.3008523088665565, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.07784366607666, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8843610286712646, "num_tokens": 390243269.0, "step": 10226 }, { "epoch": 1.300979519145147, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.057112693786621, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8686263561248779, "num_tokens": 390282488.0, "step": 10227 }, { "epoch": 1.3011067294237375, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.016844749450684, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8630608320236206, "num_tokens": 390320268.0, "step": 10228 }, { "epoch": 1.3012339397023278, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.103936195373535, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8580069541931152, "num_tokens": 390354733.0, "step": 10229 }, { "epoch": 1.3013611499809183, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.07046127319336, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8673901557922363, "num_tokens": 390388161.0, "step": 10230 }, { "epoch": 1.3014883602595089, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.041888236999512, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8769686818122864, "num_tokens": 390424609.0, "step": 10231 }, { "epoch": 1.3016155705380994, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.0130033493042, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8844008445739746, "num_tokens": 390460809.0, "step": 10232 }, { "epoch": 1.30174278081669, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.0773286819458, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8564841747283936, "num_tokens": 390500198.0, "step": 10233 }, { "epoch": 1.3018699910952805, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.039977073669434, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8608680367469788, "num_tokens": 390545588.0, "step": 10234 }, { "epoch": 1.301997201373871, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.057963371276855, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8584036827087402, "num_tokens": 390581681.0, "step": 10235 }, { "epoch": 1.3021244116524615, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.054402351379395, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8850225210189819, "num_tokens": 390619122.0, "step": 10236 }, { "epoch": 1.302251621931052, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.051960945129395, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8556299805641174, "num_tokens": 390667974.0, "step": 10237 }, { "epoch": 1.3023788322096426, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.035645484924316, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8697109222412109, "num_tokens": 390705188.0, "step": 10238 }, { "epoch": 1.302506042488233, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.050663948059082, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.876775860786438, "num_tokens": 390746892.0, "step": 10239 }, { "epoch": 1.3026332527668236, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.040348052978516, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8823236227035522, "num_tokens": 390787727.0, "step": 10240 }, { "epoch": 1.3027604630454142, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.124835968017578, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8594057559967041, "num_tokens": 390823702.0, "step": 10241 }, { "epoch": 1.3028876733240047, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.056906700134277, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.867111325263977, "num_tokens": 390860256.0, "step": 10242 }, { "epoch": 1.3030148836025952, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.05360221862793, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8809511661529541, "num_tokens": 390901676.0, "step": 10243 }, { "epoch": 1.3031420938811857, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.1032075881958, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8511795997619629, "num_tokens": 390936911.0, "step": 10244 }, { "epoch": 1.303269304159776, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.087254524230957, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8559682965278625, "num_tokens": 390976185.0, "step": 10245 }, { "epoch": 1.3033965144383666, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.071135520935059, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8676924109458923, "num_tokens": 391014661.0, "step": 10246 }, { "epoch": 1.303523724716957, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.10175895690918, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8603218793869019, "num_tokens": 391056205.0, "step": 10247 }, { "epoch": 1.3036509349955476, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.08730411529541, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8692299127578735, "num_tokens": 391092187.0, "step": 10248 }, { "epoch": 1.3037781452741382, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.049738883972168, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8828843832015991, "num_tokens": 391130890.0, "step": 10249 }, { "epoch": 1.3039053555527287, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.0425386428833, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8772585391998291, "num_tokens": 391172783.0, "step": 10250 }, { "epoch": 1.3040325658313192, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.12482738494873, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8513917922973633, "num_tokens": 391215031.0, "step": 10251 }, { "epoch": 1.3041597761099097, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.113974571228027, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8709577322006226, "num_tokens": 391251400.0, "step": 10252 }, { "epoch": 1.3042869863885003, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.128793716430664, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8587765693664551, "num_tokens": 391285895.0, "step": 10253 }, { "epoch": 1.3044141966670906, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.04940128326416, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8697693943977356, "num_tokens": 391328361.0, "step": 10254 }, { "epoch": 1.304541406945681, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.096671104431152, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8668638467788696, "num_tokens": 391365019.0, "step": 10255 }, { "epoch": 1.3046686172242716, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.15754508972168, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8517932891845703, "num_tokens": 391400407.0, "step": 10256 }, { "epoch": 1.3047958275028622, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.017187118530273, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8496427536010742, "num_tokens": 391440660.0, "step": 10257 }, { "epoch": 1.3049230377814527, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.131558418273926, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.870599627494812, "num_tokens": 391481096.0, "step": 10258 }, { "epoch": 1.3050502480600432, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.13048267364502, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8674854636192322, "num_tokens": 391516356.0, "step": 10259 }, { "epoch": 1.3051774583386337, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.047330856323242, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8565863370895386, "num_tokens": 391554691.0, "step": 10260 }, { "epoch": 1.3053046686172243, "ewc_loss": 0.02734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.076541900634766, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8530937433242798, "num_tokens": 391592473.0, "step": 10261 }, { "epoch": 1.3054318788958148, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105658531188965, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8727293014526367, "num_tokens": 391632091.0, "step": 10262 }, { "epoch": 1.3055590891744053, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.071353912353516, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8606137633323669, "num_tokens": 391676121.0, "step": 10263 }, { "epoch": 1.3056862994529959, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.033625602722168, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8651467561721802, "num_tokens": 391709664.0, "step": 10264 }, { "epoch": 1.3058135097315864, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.072007179260254, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8562666177749634, "num_tokens": 391755828.0, "step": 10265 }, { "epoch": 1.305940720010177, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.090287208557129, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8832811713218689, "num_tokens": 391800232.0, "step": 10266 }, { "epoch": 1.3060679302887674, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.100373268127441, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8686341643333435, "num_tokens": 391837612.0, "step": 10267 }, { "epoch": 1.306195140567358, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.081192970275879, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.877853274345398, "num_tokens": 391871193.0, "step": 10268 }, { "epoch": 1.3063223508459483, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.113931655883789, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8595881462097168, "num_tokens": 391908203.0, "step": 10269 }, { "epoch": 1.3064495611245388, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.103476524353027, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8731969594955444, "num_tokens": 391938744.0, "step": 10270 }, { "epoch": 1.3065767714031293, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.109128952026367, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.869684100151062, "num_tokens": 391976020.0, "step": 10271 }, { "epoch": 1.3067039816817199, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.046759605407715, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.857633113861084, "num_tokens": 392018006.0, "step": 10272 }, { "epoch": 1.3068311919603104, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.10679817199707, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8652932643890381, "num_tokens": 392055358.0, "step": 10273 }, { "epoch": 1.306958402238901, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.058182716369629, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8655713796615601, "num_tokens": 392097727.0, "step": 10274 }, { "epoch": 1.3070856125174914, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.106383323669434, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8726251721382141, "num_tokens": 392144641.0, "step": 10275 }, { "epoch": 1.307212822796082, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.080899238586426, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8750442862510681, "num_tokens": 392178810.0, "step": 10276 }, { "epoch": 1.3073400330746725, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.128405570983887, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8621395826339722, "num_tokens": 392213571.0, "step": 10277 }, { "epoch": 1.3074672433532628, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.121853828430176, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.863103449344635, "num_tokens": 392254628.0, "step": 10278 }, { "epoch": 1.3075944536318533, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.053887367248535, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8665316700935364, "num_tokens": 392297243.0, "step": 10279 }, { "epoch": 1.3077216639104439, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.227328300476074, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8695008158683777, "num_tokens": 392329116.0, "step": 10280 }, { "epoch": 1.3078488741890344, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.107409477233887, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8636165857315063, "num_tokens": 392370542.0, "step": 10281 }, { "epoch": 1.307976084467625, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.061505317687988, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8726115822792053, "num_tokens": 392411155.0, "step": 10282 }, { "epoch": 1.3081032947462155, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.19184398651123, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8675423860549927, "num_tokens": 392450440.0, "step": 10283 }, { "epoch": 1.308230505024806, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.09819221496582, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8771986961364746, "num_tokens": 392488545.0, "step": 10284 }, { "epoch": 1.3083577153033965, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.032432556152344, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.866890549659729, "num_tokens": 392527138.0, "step": 10285 }, { "epoch": 1.308484925581987, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.116230010986328, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8670626282691956, "num_tokens": 392568181.0, "step": 10286 }, { "epoch": 1.3086121358605776, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.048766136169434, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8817833065986633, "num_tokens": 392610900.0, "step": 10287 }, { "epoch": 1.308739346139168, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.084600448608398, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8587946891784668, "num_tokens": 392646624.0, "step": 10288 }, { "epoch": 1.3088665564177586, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.110179901123047, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.867996096611023, "num_tokens": 392683103.0, "step": 10289 }, { "epoch": 1.3089937666963491, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.031933784484863, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8686609864234924, "num_tokens": 392723169.0, "step": 10290 }, { "epoch": 1.3091209769749397, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.068418502807617, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8823741674423218, "num_tokens": 392762323.0, "step": 10291 }, { "epoch": 1.3092481872535302, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.099574089050293, "learning_rate": 1e-06, "loss": 0.35, "mean_token_accuracy": 0.8846819400787354, "num_tokens": 392797555.0, "step": 10292 }, { "epoch": 1.3093753975321207, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.056177139282227, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8778865933418274, "num_tokens": 392840001.0, "step": 10293 }, { "epoch": 1.309502607810711, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.0372314453125, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8673042058944702, "num_tokens": 392880871.0, "step": 10294 }, { "epoch": 1.3096298180893016, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.119302749633789, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8781142234802246, "num_tokens": 392917554.0, "step": 10295 }, { "epoch": 1.309757028367892, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.0006742477417, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8847224116325378, "num_tokens": 392958601.0, "step": 10296 }, { "epoch": 1.3098842386464826, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.076114654541016, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8698813915252686, "num_tokens": 393000108.0, "step": 10297 }, { "epoch": 1.3100114489250732, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.130208969116211, "learning_rate": 1e-06, "loss": 0.3491, "mean_token_accuracy": 0.8890432715415955, "num_tokens": 393032321.0, "step": 10298 }, { "epoch": 1.3101386592036637, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.146221160888672, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8665632009506226, "num_tokens": 393068347.0, "step": 10299 }, { "epoch": 1.3102658694822542, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.097050666809082, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8852458000183105, "num_tokens": 393102509.0, "step": 10300 }, { "epoch": 1.3103930797608447, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.108894348144531, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8764764070510864, "num_tokens": 393140974.0, "step": 10301 }, { "epoch": 1.3105202900394353, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.122133255004883, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8772505521774292, "num_tokens": 393177214.0, "step": 10302 }, { "epoch": 1.3106475003180256, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.0801362991333, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8719922304153442, "num_tokens": 393215801.0, "step": 10303 }, { "epoch": 1.310774710596616, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.102264404296875, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8795722723007202, "num_tokens": 393256444.0, "step": 10304 }, { "epoch": 1.3109019208752066, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.074822425842285, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8651798963546753, "num_tokens": 393300238.0, "step": 10305 }, { "epoch": 1.3110291311537972, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.134621620178223, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8542877435684204, "num_tokens": 393344087.0, "step": 10306 }, { "epoch": 1.3111563414323877, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.08701229095459, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8733518719673157, "num_tokens": 393381678.0, "step": 10307 }, { "epoch": 1.3112835517109782, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.149760246276855, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8668234348297119, "num_tokens": 393415740.0, "step": 10308 }, { "epoch": 1.3114107619895687, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.102877616882324, "learning_rate": 1e-06, "loss": 0.4883, "mean_token_accuracy": 0.8444958925247192, "num_tokens": 393457692.0, "step": 10309 }, { "epoch": 1.3115379722681593, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.082918167114258, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8634289503097534, "num_tokens": 393499003.0, "step": 10310 }, { "epoch": 1.3116651825467498, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.109587669372559, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8571856021881104, "num_tokens": 393533842.0, "step": 10311 }, { "epoch": 1.3117923928253403, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.10019302368164, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8758426308631897, "num_tokens": 393573906.0, "step": 10312 }, { "epoch": 1.3119196031039309, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.058201789855957, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.866591215133667, "num_tokens": 393615298.0, "step": 10313 }, { "epoch": 1.3120468133825214, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.134919166564941, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.857677161693573, "num_tokens": 393654690.0, "step": 10314 }, { "epoch": 1.312174023661112, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.103055953979492, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8702421188354492, "num_tokens": 393696222.0, "step": 10315 }, { "epoch": 1.3123012339397024, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.081323623657227, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8645203709602356, "num_tokens": 393736748.0, "step": 10316 }, { "epoch": 1.312428444218293, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.11230182647705, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8590970039367676, "num_tokens": 393773077.0, "step": 10317 }, { "epoch": 1.3125556544968833, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.085358619689941, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8793929815292358, "num_tokens": 393806466.0, "step": 10318 }, { "epoch": 1.3126828647754738, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.108469009399414, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8643585443496704, "num_tokens": 393840922.0, "step": 10319 }, { "epoch": 1.3128100750540643, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.145757675170898, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8725970387458801, "num_tokens": 393874544.0, "step": 10320 }, { "epoch": 1.3129372853326549, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.016112327575684, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8640874624252319, "num_tokens": 393917773.0, "step": 10321 }, { "epoch": 1.3130644956112454, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.07681655883789, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8637762069702148, "num_tokens": 393951525.0, "step": 10322 }, { "epoch": 1.313191705889836, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.084491729736328, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8737507462501526, "num_tokens": 393989796.0, "step": 10323 }, { "epoch": 1.3133189161684264, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.070401191711426, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8649559617042542, "num_tokens": 394028746.0, "step": 10324 }, { "epoch": 1.313446126447017, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.043133735656738, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.876710832118988, "num_tokens": 394065241.0, "step": 10325 }, { "epoch": 1.3135733367256075, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.141450881958008, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8627379536628723, "num_tokens": 394110113.0, "step": 10326 }, { "epoch": 1.3137005470041978, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.051301002502441, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.851661205291748, "num_tokens": 394146577.0, "step": 10327 }, { "epoch": 1.3138277572827883, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.068611145019531, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8677014112472534, "num_tokens": 394185392.0, "step": 10328 }, { "epoch": 1.3139549675613789, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.090361595153809, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8590327501296997, "num_tokens": 394226301.0, "step": 10329 }, { "epoch": 1.3140821778399694, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.139505386352539, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8593698740005493, "num_tokens": 394262654.0, "step": 10330 }, { "epoch": 1.31420938811856, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.100317001342773, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8797107338905334, "num_tokens": 394299185.0, "step": 10331 }, { "epoch": 1.3143365983971504, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.096404075622559, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8696715235710144, "num_tokens": 394331364.0, "step": 10332 }, { "epoch": 1.314463808675741, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.125665664672852, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8560510873794556, "num_tokens": 394375479.0, "step": 10333 }, { "epoch": 1.3145910189543315, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.122579574584961, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8742729425430298, "num_tokens": 394414913.0, "step": 10334 }, { "epoch": 1.314718229232922, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.116384506225586, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8691019415855408, "num_tokens": 394450889.0, "step": 10335 }, { "epoch": 1.3148454395115126, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.081109046936035, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8518098592758179, "num_tokens": 394495280.0, "step": 10336 }, { "epoch": 1.314972649790103, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.093803405761719, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8763395547866821, "num_tokens": 394536811.0, "step": 10337 }, { "epoch": 1.3150998600686936, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.128877639770508, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8653402328491211, "num_tokens": 394573353.0, "step": 10338 }, { "epoch": 1.3152270703472841, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.09662914276123, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8605316877365112, "num_tokens": 394603517.0, "step": 10339 }, { "epoch": 1.3153542806258747, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.118056297302246, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8736869096755981, "num_tokens": 394650547.0, "step": 10340 }, { "epoch": 1.3154814909044652, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.072230339050293, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8722884654998779, "num_tokens": 394684460.0, "step": 10341 }, { "epoch": 1.3156087011830557, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.13294506072998, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8832790851593018, "num_tokens": 394733520.0, "step": 10342 }, { "epoch": 1.315735911461646, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.088443756103516, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8618505001068115, "num_tokens": 394775498.0, "step": 10343 }, { "epoch": 1.3158631217402366, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105690956115723, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.855704665184021, "num_tokens": 394815138.0, "step": 10344 }, { "epoch": 1.315990332018827, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.159492492675781, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8597539663314819, "num_tokens": 394850447.0, "step": 10345 }, { "epoch": 1.3161175422974176, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.138643264770508, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8609525561332703, "num_tokens": 394891800.0, "step": 10346 }, { "epoch": 1.3162447525760081, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.139326095581055, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8748443722724915, "num_tokens": 394930087.0, "step": 10347 }, { "epoch": 1.3163719628545987, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.146121978759766, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8773590326309204, "num_tokens": 394964497.0, "step": 10348 }, { "epoch": 1.3164991731331892, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.130292892456055, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8502991795539856, "num_tokens": 395003127.0, "step": 10349 }, { "epoch": 1.3166263834117797, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.116178512573242, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8571653366088867, "num_tokens": 395040570.0, "step": 10350 }, { "epoch": 1.3167535936903703, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.115677833557129, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8684670925140381, "num_tokens": 395075205.0, "step": 10351 }, { "epoch": 1.3168808039689606, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.066131591796875, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8676154613494873, "num_tokens": 395117492.0, "step": 10352 }, { "epoch": 1.317008014247551, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.10003662109375, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8646436333656311, "num_tokens": 395162449.0, "step": 10353 }, { "epoch": 1.3171352245261416, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.104978561401367, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8770496845245361, "num_tokens": 395197310.0, "step": 10354 }, { "epoch": 1.3172624348047322, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.08714485168457, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8789169788360596, "num_tokens": 395232420.0, "step": 10355 }, { "epoch": 1.3173896450833227, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.138039588928223, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8474929332733154, "num_tokens": 395269392.0, "step": 10356 }, { "epoch": 1.3175168553619132, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.128012657165527, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8565693497657776, "num_tokens": 395305496.0, "step": 10357 }, { "epoch": 1.3176440656405037, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105443000793457, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8763567805290222, "num_tokens": 395347089.0, "step": 10358 }, { "epoch": 1.3177712759190943, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.093774795532227, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8782202005386353, "num_tokens": 395382885.0, "step": 10359 }, { "epoch": 1.3178984861976848, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.089652061462402, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8815767765045166, "num_tokens": 395416260.0, "step": 10360 }, { "epoch": 1.3180256964762753, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.107572555541992, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8588241338729858, "num_tokens": 395454179.0, "step": 10361 }, { "epoch": 1.3181529067548658, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.134171485900879, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8578392267227173, "num_tokens": 395489173.0, "step": 10362 }, { "epoch": 1.3182801170334564, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.13483715057373, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8618903160095215, "num_tokens": 395532057.0, "step": 10363 }, { "epoch": 1.318407327312047, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.102781295776367, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8738473653793335, "num_tokens": 395570321.0, "step": 10364 }, { "epoch": 1.3185345375906374, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.115955352783203, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8870380520820618, "num_tokens": 395609041.0, "step": 10365 }, { "epoch": 1.318661747869228, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.14614486694336, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8446542024612427, "num_tokens": 395649961.0, "step": 10366 }, { "epoch": 1.3187889581478183, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.141490936279297, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8690533638000488, "num_tokens": 395691882.0, "step": 10367 }, { "epoch": 1.3189161684264088, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.09087085723877, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8555455207824707, "num_tokens": 395731496.0, "step": 10368 }, { "epoch": 1.3190433787049993, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.077788352966309, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8714547157287598, "num_tokens": 395772618.0, "step": 10369 }, { "epoch": 1.3191705889835899, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.095870971679688, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8506076335906982, "num_tokens": 395811907.0, "step": 10370 }, { "epoch": 1.3192977992621804, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.091876029968262, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.862033486366272, "num_tokens": 395851186.0, "step": 10371 }, { "epoch": 1.319425009540771, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.126263618469238, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8673083782196045, "num_tokens": 395893868.0, "step": 10372 }, { "epoch": 1.3195522198193614, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.137646675109863, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8765450119972229, "num_tokens": 395930519.0, "step": 10373 }, { "epoch": 1.319679430097952, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.09190559387207, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8833981156349182, "num_tokens": 395965692.0, "step": 10374 }, { "epoch": 1.3198066403765425, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.142549514770508, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8693443536758423, "num_tokens": 396002316.0, "step": 10375 }, { "epoch": 1.3199338506551328, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.107414245605469, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8735814690589905, "num_tokens": 396043392.0, "step": 10376 }, { "epoch": 1.3200610609337233, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.149795532226562, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8647129535675049, "num_tokens": 396078603.0, "step": 10377 }, { "epoch": 1.3201882712123139, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.102195739746094, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8588329553604126, "num_tokens": 396121132.0, "step": 10378 }, { "epoch": 1.3203154814909044, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.128162384033203, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8768244981765747, "num_tokens": 396155730.0, "step": 10379 }, { "epoch": 1.320442691769495, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.181199073791504, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8534327745437622, "num_tokens": 396190976.0, "step": 10380 }, { "epoch": 1.3205699020480854, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.135422706604004, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8668039441108704, "num_tokens": 396229571.0, "step": 10381 }, { "epoch": 1.320697112326676, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.222223281860352, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.866763174533844, "num_tokens": 396265586.0, "step": 10382 }, { "epoch": 1.3208243226052665, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.106244087219238, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8715482354164124, "num_tokens": 396303810.0, "step": 10383 }, { "epoch": 1.320951532883857, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.104379653930664, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8872175216674805, "num_tokens": 396345671.0, "step": 10384 }, { "epoch": 1.3210787431624476, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.108417510986328, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8842103481292725, "num_tokens": 396388062.0, "step": 10385 }, { "epoch": 1.321205953441038, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.115006446838379, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8812181949615479, "num_tokens": 396425189.0, "step": 10386 }, { "epoch": 1.3213331637196286, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.128305435180664, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8579104542732239, "num_tokens": 396461861.0, "step": 10387 }, { "epoch": 1.3214603739982191, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.139511108398438, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8780019283294678, "num_tokens": 396506190.0, "step": 10388 }, { "epoch": 1.3215875842768097, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105632781982422, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8688223958015442, "num_tokens": 396538900.0, "step": 10389 }, { "epoch": 1.3217147945554002, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.122298240661621, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8768332600593567, "num_tokens": 396577029.0, "step": 10390 }, { "epoch": 1.3218420048339907, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.119379997253418, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8522220849990845, "num_tokens": 396616917.0, "step": 10391 }, { "epoch": 1.321969215112581, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.105834007263184, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8767808675765991, "num_tokens": 396651726.0, "step": 10392 }, { "epoch": 1.3220964253911716, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.158496856689453, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8556705117225647, "num_tokens": 396693796.0, "step": 10393 }, { "epoch": 1.322223635669762, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.13882064819336, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.858422040939331, "num_tokens": 396731947.0, "step": 10394 }, { "epoch": 1.3223508459483526, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 12.184514999389648, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.870768129825592, "num_tokens": 396767062.0, "step": 10395 }, { "epoch": 1.3224780562269431, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.137577056884766, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8719899654388428, "num_tokens": 396803918.0, "step": 10396 }, { "epoch": 1.3226052665055337, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.104560852050781, "learning_rate": 1e-06, "loss": 0.5006, "mean_token_accuracy": 0.8408126831054688, "num_tokens": 396850314.0, "step": 10397 }, { "epoch": 1.3227324767841242, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.087753295898438, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8613983392715454, "num_tokens": 396887920.0, "step": 10398 }, { "epoch": 1.3228596870627147, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.098979949951172, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8569913506507874, "num_tokens": 396923369.0, "step": 10399 }, { "epoch": 1.3229868973413053, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.070257186889648, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8663424253463745, "num_tokens": 396962338.0, "step": 10400 }, { "epoch": 1.3231141076198956, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.076987266540527, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8783178329467773, "num_tokens": 396998275.0, "step": 10401 }, { "epoch": 1.323241317898486, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.083297729492188, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8654285669326782, "num_tokens": 397040034.0, "step": 10402 }, { "epoch": 1.3233685281770766, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.138314247131348, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.885780394077301, "num_tokens": 397080549.0, "step": 10403 }, { "epoch": 1.3234957384556671, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.137091636657715, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8807240724563599, "num_tokens": 397120803.0, "step": 10404 }, { "epoch": 1.3236229487342577, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.138452529907227, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8722922801971436, "num_tokens": 397160463.0, "step": 10405 }, { "epoch": 1.3237501590128482, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.112936019897461, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8612892627716064, "num_tokens": 397205425.0, "step": 10406 }, { "epoch": 1.3238773692914387, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.15774154663086, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8632809519767761, "num_tokens": 397238907.0, "step": 10407 }, { "epoch": 1.3240045795700293, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.129267692565918, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.867825448513031, "num_tokens": 397281639.0, "step": 10408 }, { "epoch": 1.3241317898486198, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.132115364074707, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8560419678688049, "num_tokens": 397323350.0, "step": 10409 }, { "epoch": 1.3242590001272103, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.168781280517578, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.861353874206543, "num_tokens": 397364438.0, "step": 10410 }, { "epoch": 1.3243862104058008, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.158247947692871, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.855610728263855, "num_tokens": 397406881.0, "step": 10411 }, { "epoch": 1.3245134206843914, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.110502243041992, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8681752681732178, "num_tokens": 397451190.0, "step": 10412 }, { "epoch": 1.324640630962982, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.147315979003906, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8686895966529846, "num_tokens": 397491063.0, "step": 10413 }, { "epoch": 1.3247678412415724, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.17369556427002, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8686511516571045, "num_tokens": 397529961.0, "step": 10414 }, { "epoch": 1.324895051520163, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.100678443908691, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8780097365379333, "num_tokens": 397565616.0, "step": 10415 }, { "epoch": 1.3250222617987533, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.146132469177246, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8742746114730835, "num_tokens": 397602223.0, "step": 10416 }, { "epoch": 1.3251494720773438, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.191925048828125, "learning_rate": 1e-06, "loss": 0.4818, "mean_token_accuracy": 0.846820592880249, "num_tokens": 397638910.0, "step": 10417 }, { "epoch": 1.3252766823559343, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.104491233825684, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8600009679794312, "num_tokens": 397681200.0, "step": 10418 }, { "epoch": 1.3254038926345248, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.188685417175293, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8701666593551636, "num_tokens": 397724832.0, "step": 10419 }, { "epoch": 1.3255311029131154, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.100406646728516, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.861025333404541, "num_tokens": 397765537.0, "step": 10420 }, { "epoch": 1.325658313191706, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.139228820800781, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8593466877937317, "num_tokens": 397800945.0, "step": 10421 }, { "epoch": 1.3257855234702964, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.119622230529785, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8737632632255554, "num_tokens": 397840786.0, "step": 10422 }, { "epoch": 1.325912733748887, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.165432929992676, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8732820749282837, "num_tokens": 397875607.0, "step": 10423 }, { "epoch": 1.3260399440274775, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.11329460144043, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8834843635559082, "num_tokens": 397909524.0, "step": 10424 }, { "epoch": 1.3261671543060678, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.146409034729004, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.85196852684021, "num_tokens": 397947862.0, "step": 10425 }, { "epoch": 1.3262943645846583, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.14203929901123, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8812136054039001, "num_tokens": 397980958.0, "step": 10426 }, { "epoch": 1.3264215748632489, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.16235637664795, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8653232455253601, "num_tokens": 398021231.0, "step": 10427 }, { "epoch": 1.3265487851418394, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.099538803100586, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8659312129020691, "num_tokens": 398059202.0, "step": 10428 }, { "epoch": 1.32667599542043, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.166849136352539, "learning_rate": 1e-06, "loss": 0.3411, "mean_token_accuracy": 0.8895757794380188, "num_tokens": 398097121.0, "step": 10429 }, { "epoch": 1.3268032056990204, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.179564476013184, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8675611615180969, "num_tokens": 398130413.0, "step": 10430 }, { "epoch": 1.326930415977611, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.174125671386719, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8659896850585938, "num_tokens": 398172535.0, "step": 10431 }, { "epoch": 1.3270576262562015, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.128620147705078, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8691287040710449, "num_tokens": 398206681.0, "step": 10432 }, { "epoch": 1.327184836534792, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.163249015808105, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.871756374835968, "num_tokens": 398245206.0, "step": 10433 }, { "epoch": 1.3273120468133826, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.192840576171875, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8649284243583679, "num_tokens": 398278902.0, "step": 10434 }, { "epoch": 1.327439257091973, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.080484390258789, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.868645429611206, "num_tokens": 398315833.0, "step": 10435 }, { "epoch": 1.3275664673705636, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.108061790466309, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8503756523132324, "num_tokens": 398356496.0, "step": 10436 }, { "epoch": 1.3276936776491541, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.116205215454102, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8740681409835815, "num_tokens": 398393661.0, "step": 10437 }, { "epoch": 1.3278208879277447, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.116966247558594, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8846657872200012, "num_tokens": 398430463.0, "step": 10438 }, { "epoch": 1.3279480982063352, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.120992660522461, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8662765622138977, "num_tokens": 398470382.0, "step": 10439 }, { "epoch": 1.3280753084849257, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.194063186645508, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8504562973976135, "num_tokens": 398504793.0, "step": 10440 }, { "epoch": 1.328202518763516, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.137944221496582, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8588345646858215, "num_tokens": 398541038.0, "step": 10441 }, { "epoch": 1.3283297290421066, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.098916053771973, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8507903814315796, "num_tokens": 398578047.0, "step": 10442 }, { "epoch": 1.328456939320697, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.134302139282227, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8572660684585571, "num_tokens": 398618946.0, "step": 10443 }, { "epoch": 1.3285841495992876, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.105095863342285, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8728563189506531, "num_tokens": 398656895.0, "step": 10444 }, { "epoch": 1.3287113598778781, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.13746452331543, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8683537840843201, "num_tokens": 398696521.0, "step": 10445 }, { "epoch": 1.3288385701564687, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.100339889526367, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8690739274024963, "num_tokens": 398735275.0, "step": 10446 }, { "epoch": 1.3289657804350592, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.133293151855469, "learning_rate": 1e-06, "loss": 0.4645, "mean_token_accuracy": 0.8528019189834595, "num_tokens": 398774753.0, "step": 10447 }, { "epoch": 1.3290929907136497, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.164984703063965, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8611285090446472, "num_tokens": 398810893.0, "step": 10448 }, { "epoch": 1.3292202009922403, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.160447120666504, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.862133264541626, "num_tokens": 398849741.0, "step": 10449 }, { "epoch": 1.3293474112708306, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.200562477111816, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8622243404388428, "num_tokens": 398881186.0, "step": 10450 }, { "epoch": 1.329474621549421, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.117599487304688, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.884790301322937, "num_tokens": 398918677.0, "step": 10451 }, { "epoch": 1.3296018318280116, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.129297256469727, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8645374774932861, "num_tokens": 398953435.0, "step": 10452 }, { "epoch": 1.3297290421066021, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.16401195526123, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8748147487640381, "num_tokens": 398991799.0, "step": 10453 }, { "epoch": 1.3298562523851927, "ewc_loss": 0.027587890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.13453197479248, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8725548982620239, "num_tokens": 399023211.0, "step": 10454 }, { "epoch": 1.3299834626637832, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.174210548400879, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8643463253974915, "num_tokens": 399061431.0, "step": 10455 }, { "epoch": 1.3301106729423737, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.149880409240723, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8691913485527039, "num_tokens": 399097384.0, "step": 10456 }, { "epoch": 1.3302378832209643, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.186251640319824, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.87034010887146, "num_tokens": 399133935.0, "step": 10457 }, { "epoch": 1.3303650934995548, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.128731727600098, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8725641965866089, "num_tokens": 399165709.0, "step": 10458 }, { "epoch": 1.3304923037781453, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.160801887512207, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8658801317214966, "num_tokens": 399203548.0, "step": 10459 }, { "epoch": 1.3306195140567358, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.136713981628418, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8660668730735779, "num_tokens": 399244341.0, "step": 10460 }, { "epoch": 1.3307467243353264, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.098298072814941, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8727098703384399, "num_tokens": 399282284.0, "step": 10461 }, { "epoch": 1.330873934613917, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.125475883483887, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8805162906646729, "num_tokens": 399320523.0, "step": 10462 }, { "epoch": 1.3310011448925074, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.104510307312012, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8684684634208679, "num_tokens": 399357489.0, "step": 10463 }, { "epoch": 1.331128355171098, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.142547607421875, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8557853698730469, "num_tokens": 399393561.0, "step": 10464 }, { "epoch": 1.3312555654496883, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.12877368927002, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8741384148597717, "num_tokens": 399424514.0, "step": 10465 }, { "epoch": 1.3313827757282788, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.108098983764648, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8731844425201416, "num_tokens": 399461372.0, "step": 10466 }, { "epoch": 1.3315099860068693, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.113636016845703, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8816266059875488, "num_tokens": 399502503.0, "step": 10467 }, { "epoch": 1.3316371962854598, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.141709327697754, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8619435429573059, "num_tokens": 399537836.0, "step": 10468 }, { "epoch": 1.3317644065640504, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.132166862487793, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8630883693695068, "num_tokens": 399575189.0, "step": 10469 }, { "epoch": 1.331891616842641, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.272808074951172, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8493124842643738, "num_tokens": 399614406.0, "step": 10470 }, { "epoch": 1.3320188271212314, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.101582527160645, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8728863000869751, "num_tokens": 399652847.0, "step": 10471 }, { "epoch": 1.332146037399822, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.14393138885498, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8577035665512085, "num_tokens": 399691628.0, "step": 10472 }, { "epoch": 1.3322732476784125, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.114538192749023, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8788655400276184, "num_tokens": 399732875.0, "step": 10473 }, { "epoch": 1.3324004579570028, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.167851448059082, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8782903552055359, "num_tokens": 399769537.0, "step": 10474 }, { "epoch": 1.3325276682355933, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.125940322875977, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8514247536659241, "num_tokens": 399807757.0, "step": 10475 }, { "epoch": 1.3326548785141838, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.194887161254883, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8745262622833252, "num_tokens": 399837351.0, "step": 10476 }, { "epoch": 1.3327820887927744, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.091545104980469, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8861836791038513, "num_tokens": 399873678.0, "step": 10477 }, { "epoch": 1.332909299071365, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.140889167785645, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.869369387626648, "num_tokens": 399916206.0, "step": 10478 }, { "epoch": 1.3330365093499554, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.130671501159668, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8798829317092896, "num_tokens": 399954144.0, "step": 10479 }, { "epoch": 1.333163719628546, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.086322784423828, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8758354783058167, "num_tokens": 399995865.0, "step": 10480 }, { "epoch": 1.3332909299071365, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.178533554077148, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8692915439605713, "num_tokens": 400033234.0, "step": 10481 }, { "epoch": 1.333418140185727, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.143486976623535, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8769842386245728, "num_tokens": 400066842.0, "step": 10482 }, { "epoch": 1.3335453504643175, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.145618438720703, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8798189163208008, "num_tokens": 400099380.0, "step": 10483 }, { "epoch": 1.333672560742908, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.108022689819336, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.879946768283844, "num_tokens": 400137419.0, "step": 10484 }, { "epoch": 1.3337997710214986, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.124211311340332, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8689180016517639, "num_tokens": 400171086.0, "step": 10485 }, { "epoch": 1.3339269813000891, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.122798919677734, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8665903806686401, "num_tokens": 400206763.0, "step": 10486 }, { "epoch": 1.3340541915786797, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.107308387756348, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8529538512229919, "num_tokens": 400242708.0, "step": 10487 }, { "epoch": 1.3341814018572702, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.121993064880371, "learning_rate": 1e-06, "loss": 0.4715, "mean_token_accuracy": 0.8484469652175903, "num_tokens": 400277417.0, "step": 10488 }, { "epoch": 1.3343086121358605, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.171826362609863, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.868161678314209, "num_tokens": 400315811.0, "step": 10489 }, { "epoch": 1.334435822414451, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.06775188446045, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8491478562355042, "num_tokens": 400356321.0, "step": 10490 }, { "epoch": 1.3345630326930416, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.15328598022461, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.86849045753479, "num_tokens": 400392382.0, "step": 10491 }, { "epoch": 1.334690242971632, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.105267524719238, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8751846551895142, "num_tokens": 400434947.0, "step": 10492 }, { "epoch": 1.3348174532502226, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.106915473937988, "learning_rate": 1e-06, "loss": 0.4664, "mean_token_accuracy": 0.8492463827133179, "num_tokens": 400472742.0, "step": 10493 }, { "epoch": 1.3349446635288131, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.134673118591309, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8816112279891968, "num_tokens": 400510029.0, "step": 10494 }, { "epoch": 1.3350718738074037, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.094917297363281, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8647072315216064, "num_tokens": 400547383.0, "step": 10495 }, { "epoch": 1.3351990840859942, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.089132308959961, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8785498738288879, "num_tokens": 400585242.0, "step": 10496 }, { "epoch": 1.3353262943645847, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.084100723266602, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8798420429229736, "num_tokens": 400623117.0, "step": 10497 }, { "epoch": 1.3354535046431752, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.083861351013184, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8575680255889893, "num_tokens": 400660170.0, "step": 10498 }, { "epoch": 1.3355807149217656, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.104891777038574, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8745141625404358, "num_tokens": 400700467.0, "step": 10499 }, { "epoch": 1.335707925200356, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.0960111618042, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8621456027030945, "num_tokens": 400740269.0, "step": 10500 }, { "epoch": 1.3358351354789466, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.11602783203125, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.878153383731842, "num_tokens": 400776465.0, "step": 10501 }, { "epoch": 1.3359623457575371, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.079020500183105, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8659317493438721, "num_tokens": 400811210.0, "step": 10502 }, { "epoch": 1.3360895560361277, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.078079223632812, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8610964417457581, "num_tokens": 400848820.0, "step": 10503 }, { "epoch": 1.3362167663147182, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.120237350463867, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8698217272758484, "num_tokens": 400888774.0, "step": 10504 }, { "epoch": 1.3363439765933087, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.102971076965332, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8706331253051758, "num_tokens": 400935067.0, "step": 10505 }, { "epoch": 1.3364711868718993, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.163326263427734, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8772803544998169, "num_tokens": 400974122.0, "step": 10506 }, { "epoch": 1.3365983971504898, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.114065170288086, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.877515971660614, "num_tokens": 401008911.0, "step": 10507 }, { "epoch": 1.3367256074290803, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.175995826721191, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8749047517776489, "num_tokens": 401039453.0, "step": 10508 }, { "epoch": 1.3368528177076708, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.104313850402832, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8777399659156799, "num_tokens": 401075854.0, "step": 10509 }, { "epoch": 1.3369800279862614, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.177703857421875, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8669262528419495, "num_tokens": 401114922.0, "step": 10510 }, { "epoch": 1.337107238264852, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.143542289733887, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8769400715827942, "num_tokens": 401155833.0, "step": 10511 }, { "epoch": 1.3372344485434424, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.13503646850586, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8740600347518921, "num_tokens": 401196971.0, "step": 10512 }, { "epoch": 1.337361658822033, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.13413143157959, "learning_rate": 1e-06, "loss": 0.5098, "mean_token_accuracy": 0.8373091220855713, "num_tokens": 401237312.0, "step": 10513 }, { "epoch": 1.3374888691006233, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.149232864379883, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.865551233291626, "num_tokens": 401274630.0, "step": 10514 }, { "epoch": 1.3376160793792138, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.140791893005371, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8731082677841187, "num_tokens": 401311780.0, "step": 10515 }, { "epoch": 1.3377432896578043, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.134784698486328, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8767846822738647, "num_tokens": 401352614.0, "step": 10516 }, { "epoch": 1.3378704999363948, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.105480194091797, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8715455532073975, "num_tokens": 401394495.0, "step": 10517 }, { "epoch": 1.3379977102149854, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.191664695739746, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8549661636352539, "num_tokens": 401427758.0, "step": 10518 }, { "epoch": 1.338124920493576, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.161428451538086, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8536293506622314, "num_tokens": 401473868.0, "step": 10519 }, { "epoch": 1.3382521307721664, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.10978889465332, "learning_rate": 1e-06, "loss": 0.4724, "mean_token_accuracy": 0.8513860702514648, "num_tokens": 401510459.0, "step": 10520 }, { "epoch": 1.338379341050757, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.237381935119629, "learning_rate": 1e-06, "loss": 0.508, "mean_token_accuracy": 0.8409914970397949, "num_tokens": 401545283.0, "step": 10521 }, { "epoch": 1.3385065513293475, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.199090957641602, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8691912889480591, "num_tokens": 401577166.0, "step": 10522 }, { "epoch": 1.3386337616079378, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.108291625976562, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8812912106513977, "num_tokens": 401619075.0, "step": 10523 }, { "epoch": 1.3387609718865283, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.157631874084473, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8622995615005493, "num_tokens": 401661286.0, "step": 10524 }, { "epoch": 1.3388881821651188, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.168333053588867, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8591436147689819, "num_tokens": 401699672.0, "step": 10525 }, { "epoch": 1.3390153924437094, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.157596588134766, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8625928163528442, "num_tokens": 401732676.0, "step": 10526 }, { "epoch": 1.3391426027223, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.217352867126465, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.866669237613678, "num_tokens": 401771593.0, "step": 10527 }, { "epoch": 1.3392698130008904, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.11944580078125, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.855059027671814, "num_tokens": 401809605.0, "step": 10528 }, { "epoch": 1.339397023279481, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.122414588928223, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8591955900192261, "num_tokens": 401845349.0, "step": 10529 }, { "epoch": 1.3395242335580715, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.148534774780273, "learning_rate": 1e-06, "loss": 0.3313, "mean_token_accuracy": 0.8912943005561829, "num_tokens": 401883186.0, "step": 10530 }, { "epoch": 1.339651443836662, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.178406715393066, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8522895574569702, "num_tokens": 401921149.0, "step": 10531 }, { "epoch": 1.3397786541152525, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.135154724121094, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8611432909965515, "num_tokens": 401961813.0, "step": 10532 }, { "epoch": 1.339905864393843, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.174459457397461, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8509489893913269, "num_tokens": 402001212.0, "step": 10533 }, { "epoch": 1.3400330746724336, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.162531852722168, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8601917028427124, "num_tokens": 402044648.0, "step": 10534 }, { "epoch": 1.3401602849510241, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.109014511108398, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8709723949432373, "num_tokens": 402087391.0, "step": 10535 }, { "epoch": 1.3402874952296147, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.169519424438477, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8874863386154175, "num_tokens": 402119598.0, "step": 10536 }, { "epoch": 1.3404147055082052, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.113714218139648, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8435194492340088, "num_tokens": 402153628.0, "step": 10537 }, { "epoch": 1.3405419157867955, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.155245780944824, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8840447664260864, "num_tokens": 402190876.0, "step": 10538 }, { "epoch": 1.340669126065386, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.097833633422852, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8766797184944153, "num_tokens": 402227168.0, "step": 10539 }, { "epoch": 1.3407963363439765, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.18763256072998, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8767512440681458, "num_tokens": 402262731.0, "step": 10540 }, { "epoch": 1.340923546622567, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.172325134277344, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8691881895065308, "num_tokens": 402299992.0, "step": 10541 }, { "epoch": 1.3410507569011576, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.158530235290527, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8609007000923157, "num_tokens": 402339129.0, "step": 10542 }, { "epoch": 1.3411779671797481, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.149712562561035, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8517266511917114, "num_tokens": 402375380.0, "step": 10543 }, { "epoch": 1.3413051774583387, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.129518508911133, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8561255931854248, "num_tokens": 402412159.0, "step": 10544 }, { "epoch": 1.3414323877369292, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.146642684936523, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8641109466552734, "num_tokens": 402446676.0, "step": 10545 }, { "epoch": 1.3415595980155197, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.164870262145996, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8683120012283325, "num_tokens": 402490741.0, "step": 10546 }, { "epoch": 1.3416868082941102, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.164938926696777, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8763889670372009, "num_tokens": 402527398.0, "step": 10547 }, { "epoch": 1.3418140185727006, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.150564193725586, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8674696683883667, "num_tokens": 402565914.0, "step": 10548 }, { "epoch": 1.341941228851291, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.206308364868164, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8712157011032104, "num_tokens": 402600671.0, "step": 10549 }, { "epoch": 1.3420684391298816, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.09057903289795, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8809152841567993, "num_tokens": 402637381.0, "step": 10550 }, { "epoch": 1.3421956494084721, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.217962265014648, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8721488118171692, "num_tokens": 402671190.0, "step": 10551 }, { "epoch": 1.3423228596870627, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.097674369812012, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8679089546203613, "num_tokens": 402721518.0, "step": 10552 }, { "epoch": 1.3424500699656532, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.165452003479004, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8688997626304626, "num_tokens": 402758949.0, "step": 10553 }, { "epoch": 1.3425772802442437, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.128595352172852, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8628498315811157, "num_tokens": 402794342.0, "step": 10554 }, { "epoch": 1.3427044905228342, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.217544555664062, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8792979717254639, "num_tokens": 402831031.0, "step": 10555 }, { "epoch": 1.3428317008014248, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.102949142456055, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8787456154823303, "num_tokens": 402871486.0, "step": 10556 }, { "epoch": 1.3429589110800153, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.113372802734375, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8808480501174927, "num_tokens": 402916024.0, "step": 10557 }, { "epoch": 1.3430861213586058, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.149948120117188, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8501449823379517, "num_tokens": 402952116.0, "step": 10558 }, { "epoch": 1.3432133316371964, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.147260665893555, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8666781783103943, "num_tokens": 402998064.0, "step": 10559 }, { "epoch": 1.343340541915787, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.130610466003418, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8732807636260986, "num_tokens": 403041763.0, "step": 10560 }, { "epoch": 1.3434677521943774, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.138541221618652, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8795726299285889, "num_tokens": 403080956.0, "step": 10561 }, { "epoch": 1.343594962472968, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.1843843460083, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8775573372840881, "num_tokens": 403117200.0, "step": 10562 }, { "epoch": 1.3437221727515583, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.176989555358887, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8681200742721558, "num_tokens": 403154112.0, "step": 10563 }, { "epoch": 1.3438493830301488, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.142048835754395, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8660489916801453, "num_tokens": 403188265.0, "step": 10564 }, { "epoch": 1.3439765933087393, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.124381065368652, "learning_rate": 1e-06, "loss": 0.4737, "mean_token_accuracy": 0.8463426828384399, "num_tokens": 403231236.0, "step": 10565 }, { "epoch": 1.3441038035873298, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.12529468536377, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8701385259628296, "num_tokens": 403268642.0, "step": 10566 }, { "epoch": 1.3442310138659204, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.169512748718262, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8760042190551758, "num_tokens": 403307679.0, "step": 10567 }, { "epoch": 1.344358224144511, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.080496788024902, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8747880458831787, "num_tokens": 403346395.0, "step": 10568 }, { "epoch": 1.3444854344231014, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.139389991760254, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.877511739730835, "num_tokens": 403383001.0, "step": 10569 }, { "epoch": 1.344612644701692, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.132610321044922, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8612044453620911, "num_tokens": 403426090.0, "step": 10570 }, { "epoch": 1.3447398549802825, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.132237434387207, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8616774082183838, "num_tokens": 403470231.0, "step": 10571 }, { "epoch": 1.3448670652588728, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.157031059265137, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8736668825149536, "num_tokens": 403508677.0, "step": 10572 }, { "epoch": 1.3449942755374633, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.158260345458984, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8486764430999756, "num_tokens": 403545175.0, "step": 10573 }, { "epoch": 1.3451214858160538, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.152865409851074, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8818091154098511, "num_tokens": 403583321.0, "step": 10574 }, { "epoch": 1.3452486960946444, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.181496620178223, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8619016408920288, "num_tokens": 403618547.0, "step": 10575 }, { "epoch": 1.345375906373235, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.178781509399414, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8714672923088074, "num_tokens": 403657552.0, "step": 10576 }, { "epoch": 1.3455031166518254, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.132448196411133, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8711314797401428, "num_tokens": 403697021.0, "step": 10577 }, { "epoch": 1.345630326930416, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.163372993469238, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8580924272537231, "num_tokens": 403736480.0, "step": 10578 }, { "epoch": 1.3457575372090065, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.180188179016113, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8750113248825073, "num_tokens": 403772056.0, "step": 10579 }, { "epoch": 1.345884747487597, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.145140647888184, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8627762794494629, "num_tokens": 403813646.0, "step": 10580 }, { "epoch": 1.3460119577661875, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.147323608398438, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8714444041252136, "num_tokens": 403853904.0, "step": 10581 }, { "epoch": 1.346139168044778, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.170049667358398, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8524012565612793, "num_tokens": 403898528.0, "step": 10582 }, { "epoch": 1.3462663783233686, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.219855308532715, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8534349203109741, "num_tokens": 403929518.0, "step": 10583 }, { "epoch": 1.3463935886019591, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.129720687866211, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8701299428939819, "num_tokens": 403970323.0, "step": 10584 }, { "epoch": 1.3465207988805497, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.197675704956055, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8714043498039246, "num_tokens": 404012499.0, "step": 10585 }, { "epoch": 1.3466480091591402, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.19234561920166, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8640536069869995, "num_tokens": 404055370.0, "step": 10586 }, { "epoch": 1.3467752194377305, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.182524681091309, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8686699867248535, "num_tokens": 404092498.0, "step": 10587 }, { "epoch": 1.346902429716321, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.208131790161133, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8744499683380127, "num_tokens": 404132466.0, "step": 10588 }, { "epoch": 1.3470296399949115, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.160879135131836, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8682197332382202, "num_tokens": 404171913.0, "step": 10589 }, { "epoch": 1.347156850273502, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.23099136352539, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8816964626312256, "num_tokens": 404205817.0, "step": 10590 }, { "epoch": 1.3472840605520926, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.128231048583984, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8716368675231934, "num_tokens": 404243922.0, "step": 10591 }, { "epoch": 1.3474112708306831, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.211653709411621, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8713133931159973, "num_tokens": 404283392.0, "step": 10592 }, { "epoch": 1.3475384811092737, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.205652236938477, "learning_rate": 1e-06, "loss": 0.49, "mean_token_accuracy": 0.8426167368888855, "num_tokens": 404323044.0, "step": 10593 }, { "epoch": 1.3476656913878642, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.205337524414062, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8750509023666382, "num_tokens": 404359761.0, "step": 10594 }, { "epoch": 1.3477929016664547, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.170004844665527, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8557724356651306, "num_tokens": 404395354.0, "step": 10595 }, { "epoch": 1.3479201119450452, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.190278053283691, "learning_rate": 1e-06, "loss": 0.4817, "mean_token_accuracy": 0.8525955080986023, "num_tokens": 404432052.0, "step": 10596 }, { "epoch": 1.3480473222236355, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.15797233581543, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8607019186019897, "num_tokens": 404472689.0, "step": 10597 }, { "epoch": 1.348174532502226, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.148802757263184, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8535541296005249, "num_tokens": 404515916.0, "step": 10598 }, { "epoch": 1.3483017427808166, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.202547073364258, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8672831058502197, "num_tokens": 404557399.0, "step": 10599 }, { "epoch": 1.3484289530594071, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.22831916809082, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8649293780326843, "num_tokens": 404592573.0, "step": 10600 }, { "epoch": 1.3485561633379977, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.152215957641602, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8647741079330444, "num_tokens": 404635266.0, "step": 10601 }, { "epoch": 1.3486833736165882, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.18632984161377, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8771443963050842, "num_tokens": 404673004.0, "step": 10602 }, { "epoch": 1.3488105838951787, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.223822593688965, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8526884317398071, "num_tokens": 404705153.0, "step": 10603 }, { "epoch": 1.3489377941737692, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.241290092468262, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8591939210891724, "num_tokens": 404737068.0, "step": 10604 }, { "epoch": 1.3490650044523598, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.169440269470215, "learning_rate": 1e-06, "loss": 0.4873, "mean_token_accuracy": 0.8471670150756836, "num_tokens": 404773680.0, "step": 10605 }, { "epoch": 1.3491922147309503, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.216444969177246, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8741058111190796, "num_tokens": 404809937.0, "step": 10606 }, { "epoch": 1.3493194250095408, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.147295951843262, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8590837717056274, "num_tokens": 404847146.0, "step": 10607 }, { "epoch": 1.3494466352881314, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.144213676452637, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8688874840736389, "num_tokens": 404882952.0, "step": 10608 }, { "epoch": 1.3495738455667219, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.190670013427734, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8773527145385742, "num_tokens": 404925347.0, "step": 10609 }, { "epoch": 1.3497010558453124, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.112793922424316, "learning_rate": 1e-06, "loss": 0.4667, "mean_token_accuracy": 0.8529551029205322, "num_tokens": 404963411.0, "step": 10610 }, { "epoch": 1.349828266123903, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.191603660583496, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8666926622390747, "num_tokens": 404997522.0, "step": 10611 }, { "epoch": 1.3499554764024932, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.17231273651123, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8665416240692139, "num_tokens": 405034049.0, "step": 10612 }, { "epoch": 1.3500826866810838, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.14122200012207, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8646533489227295, "num_tokens": 405074985.0, "step": 10613 }, { "epoch": 1.3502098969596743, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.136514663696289, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8822701573371887, "num_tokens": 405116133.0, "step": 10614 }, { "epoch": 1.3503371072382648, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.166678428649902, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8657565116882324, "num_tokens": 405155572.0, "step": 10615 }, { "epoch": 1.3504643175168554, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.140029907226562, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8697649240493774, "num_tokens": 405194700.0, "step": 10616 }, { "epoch": 1.350591527795446, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.169692039489746, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8631510138511658, "num_tokens": 405233582.0, "step": 10617 }, { "epoch": 1.3507187380740364, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.11515998840332, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8796879053115845, "num_tokens": 405274987.0, "step": 10618 }, { "epoch": 1.350845948352627, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.200418472290039, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8650945425033569, "num_tokens": 405308151.0, "step": 10619 }, { "epoch": 1.3509731586312175, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.126078605651855, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8677168488502502, "num_tokens": 405343939.0, "step": 10620 }, { "epoch": 1.3511003689098078, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.181385040283203, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8667244911193848, "num_tokens": 405384211.0, "step": 10621 }, { "epoch": 1.3512275791883983, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.18284797668457, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8747878074645996, "num_tokens": 405419971.0, "step": 10622 }, { "epoch": 1.3513547894669888, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.175375938415527, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8611101508140564, "num_tokens": 405451322.0, "step": 10623 }, { "epoch": 1.3514819997455794, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.14898681640625, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8592150211334229, "num_tokens": 405496101.0, "step": 10624 }, { "epoch": 1.35160921002417, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.177035331726074, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8514727354049683, "num_tokens": 405530034.0, "step": 10625 }, { "epoch": 1.3517364203027604, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.16704273223877, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8787078857421875, "num_tokens": 405570893.0, "step": 10626 }, { "epoch": 1.351863630581351, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.086743354797363, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8549916744232178, "num_tokens": 405610120.0, "step": 10627 }, { "epoch": 1.3519908408599415, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.169149398803711, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8670523166656494, "num_tokens": 405651878.0, "step": 10628 }, { "epoch": 1.352118051138532, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.137654304504395, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8722835779190063, "num_tokens": 405694578.0, "step": 10629 }, { "epoch": 1.3522452614171225, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.15750789642334, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8733077049255371, "num_tokens": 405734128.0, "step": 10630 }, { "epoch": 1.352372471695713, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.178462028503418, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8733245134353638, "num_tokens": 405776588.0, "step": 10631 }, { "epoch": 1.3524996819743036, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.170463562011719, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8626446723937988, "num_tokens": 405810867.0, "step": 10632 }, { "epoch": 1.3526268922528941, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.125518798828125, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8855471014976501, "num_tokens": 405848884.0, "step": 10633 }, { "epoch": 1.3527541025314846, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.180859565734863, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.886195719242096, "num_tokens": 405888759.0, "step": 10634 }, { "epoch": 1.3528813128100752, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.114705085754395, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8803881406784058, "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": 12.1756010055542, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8805010318756104, "num_tokens": 405960836.0, "step": 10636 }, { "epoch": 1.353135733367256, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.119036674499512, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8782031536102295, "num_tokens": 406000314.0, "step": 10637 }, { "epoch": 1.3532629436458465, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.20427131652832, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8775397539138794, "num_tokens": 406039468.0, "step": 10638 }, { "epoch": 1.353390153924437, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.156068801879883, "learning_rate": 1e-06, "loss": 0.3425, "mean_token_accuracy": 0.8921689391136169, "num_tokens": 406076828.0, "step": 10639 }, { "epoch": 1.3535173642030276, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.239434242248535, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8666956424713135, "num_tokens": 406116504.0, "step": 10640 }, { "epoch": 1.3536445744816181, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.164977073669434, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8675388693809509, "num_tokens": 406155184.0, "step": 10641 }, { "epoch": 1.3537717847602087, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.258630752563477, "learning_rate": 1e-06, "loss": 0.4915, "mean_token_accuracy": 0.8492880463600159, "num_tokens": 406195411.0, "step": 10642 }, { "epoch": 1.3538989950387992, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.209992408752441, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8584488034248352, "num_tokens": 406230159.0, "step": 10643 }, { "epoch": 1.3540262053173897, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.225471496582031, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8572472333908081, "num_tokens": 406264541.0, "step": 10644 }, { "epoch": 1.3541534155959802, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.214640617370605, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8738858103752136, "num_tokens": 406301869.0, "step": 10645 }, { "epoch": 1.3542806258745705, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.244150161743164, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8697327971458435, "num_tokens": 406338819.0, "step": 10646 }, { "epoch": 1.354407836153161, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.17105484008789, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8694865107536316, "num_tokens": 406375914.0, "step": 10647 }, { "epoch": 1.3545350464317516, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.17406177520752, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.873354971408844, "num_tokens": 406418543.0, "step": 10648 }, { "epoch": 1.3546622567103421, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.181390762329102, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8666465282440186, "num_tokens": 406455731.0, "step": 10649 }, { "epoch": 1.3547894669889327, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.20914363861084, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8675727844238281, "num_tokens": 406496266.0, "step": 10650 }, { "epoch": 1.3549166772675232, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.145421028137207, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8629797697067261, "num_tokens": 406541676.0, "step": 10651 }, { "epoch": 1.3550438875461137, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.16837215423584, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.864517092704773, "num_tokens": 406578997.0, "step": 10652 }, { "epoch": 1.3551710978247042, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.1997709274292, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8731065988540649, "num_tokens": 406619404.0, "step": 10653 }, { "epoch": 1.3552983081032948, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.17839241027832, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.8850011825561523, "num_tokens": 406656092.0, "step": 10654 }, { "epoch": 1.3554255183818853, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.159567832946777, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8511450290679932, "num_tokens": 406701600.0, "step": 10655 }, { "epoch": 1.3555527286604758, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.821487426757812e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.16642951965332, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8671324849128723, "num_tokens": 406739649.0, "step": 10656 }, { "epoch": 1.3556799389390664, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.159493446350098, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.8873963356018066, "num_tokens": 406774962.0, "step": 10657 }, { "epoch": 1.3558071492176569, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.136765480041504, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8654986023902893, "num_tokens": 406818047.0, "step": 10658 }, { "epoch": 1.3559343594962474, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.1685209274292, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8613394498825073, "num_tokens": 406859464.0, "step": 10659 }, { "epoch": 1.356061569774838, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.16380500793457, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8666408061981201, "num_tokens": 406896444.0, "step": 10660 }, { "epoch": 1.3561887800534282, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.217686653137207, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8627777695655823, "num_tokens": 406934623.0, "step": 10661 }, { "epoch": 1.3563159903320188, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.143404960632324, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.860785961151123, "num_tokens": 406974290.0, "step": 10662 }, { "epoch": 1.3564432006106093, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.163755416870117, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.886900782585144, "num_tokens": 407011279.0, "step": 10663 }, { "epoch": 1.3565704108891998, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.144071578979492, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8573974370956421, "num_tokens": 407054733.0, "step": 10664 }, { "epoch": 1.3566976211677904, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.163993835449219, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8688358664512634, "num_tokens": 407098410.0, "step": 10665 }, { "epoch": 1.3568248314463809, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.184049606323242, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8623946905136108, "num_tokens": 407134231.0, "step": 10666 }, { "epoch": 1.3569520417249714, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.16392993927002, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8539109230041504, "num_tokens": 407167231.0, "step": 10667 }, { "epoch": 1.357079252003562, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.158061027526855, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.870792031288147, "num_tokens": 407206604.0, "step": 10668 }, { "epoch": 1.3572064622821525, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.11861515045166, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8709321022033691, "num_tokens": 407248247.0, "step": 10669 }, { "epoch": 1.3573336725607428, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.180015563964844, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8776373863220215, "num_tokens": 407282870.0, "step": 10670 }, { "epoch": 1.3574608828393333, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.135095596313477, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8743293285369873, "num_tokens": 407318704.0, "step": 10671 }, { "epoch": 1.3575880931179238, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.219488143920898, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8703703880310059, "num_tokens": 407360003.0, "step": 10672 }, { "epoch": 1.3577153033965144, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.138318061828613, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8623712658882141, "num_tokens": 407399473.0, "step": 10673 }, { "epoch": 1.357842513675105, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.19823932647705, "learning_rate": 1e-06, "loss": 0.5046, "mean_token_accuracy": 0.8377596139907837, "num_tokens": 407440576.0, "step": 10674 }, { "epoch": 1.3579697239536954, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.162856101989746, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8644425868988037, "num_tokens": 407481057.0, "step": 10675 }, { "epoch": 1.358096934232286, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.162150382995605, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8801095485687256, "num_tokens": 407517104.0, "step": 10676 }, { "epoch": 1.3582241445108765, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.170150756835938, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8757514953613281, "num_tokens": 407552604.0, "step": 10677 }, { "epoch": 1.358351354789467, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.181599617004395, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8823609352111816, "num_tokens": 407588282.0, "step": 10678 }, { "epoch": 1.3584785650680575, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.196319580078125, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8640763163566589, "num_tokens": 407619837.0, "step": 10679 }, { "epoch": 1.358605775346648, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.154345512390137, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.865551233291626, "num_tokens": 407656939.0, "step": 10680 }, { "epoch": 1.3587329856252386, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.190011024475098, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8584944605827332, "num_tokens": 407702756.0, "step": 10681 }, { "epoch": 1.3588601959038291, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.171638488769531, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8767738938331604, "num_tokens": 407736142.0, "step": 10682 }, { "epoch": 1.3589874061824196, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.216126441955566, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8651981949806213, "num_tokens": 407773184.0, "step": 10683 }, { "epoch": 1.3591146164610102, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.239970207214355, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8536827564239502, "num_tokens": 407806543.0, "step": 10684 }, { "epoch": 1.3592418267396005, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.215229034423828, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8674917817115784, "num_tokens": 407841371.0, "step": 10685 }, { "epoch": 1.359369037018191, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.209489822387695, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8713631629943848, "num_tokens": 407879862.0, "step": 10686 }, { "epoch": 1.3594962472967815, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.168070793151855, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8616953492164612, "num_tokens": 407917947.0, "step": 10687 }, { "epoch": 1.359623457575372, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.210240364074707, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8758754730224609, "num_tokens": 407953987.0, "step": 10688 }, { "epoch": 1.3597506678539626, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.176817893981934, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8520671129226685, "num_tokens": 407993166.0, "step": 10689 }, { "epoch": 1.3598778781325531, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.143589973449707, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8766376972198486, "num_tokens": 408037063.0, "step": 10690 }, { "epoch": 1.3600050884111436, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.207818984985352, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8707046508789062, "num_tokens": 408078526.0, "step": 10691 }, { "epoch": 1.3601322986897342, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.13888168334961, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.875556230545044, "num_tokens": 408116002.0, "step": 10692 }, { "epoch": 1.3602595089683247, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.157674789428711, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8643594980239868, "num_tokens": 408155885.0, "step": 10693 }, { "epoch": 1.3603867192469152, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.261946678161621, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8731585144996643, "num_tokens": 408190085.0, "step": 10694 }, { "epoch": 1.3605139295255055, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.181649208068848, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8698375225067139, "num_tokens": 408227289.0, "step": 10695 }, { "epoch": 1.360641139804096, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.15603256225586, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8761791586875916, "num_tokens": 408271853.0, "step": 10696 }, { "epoch": 1.3607683500826866, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.178955078125, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8806302547454834, "num_tokens": 408311133.0, "step": 10697 }, { "epoch": 1.3608955603612771, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.188179969787598, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8702564835548401, "num_tokens": 408342775.0, "step": 10698 }, { "epoch": 1.3610227706398677, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.226613998413086, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8804448246955872, "num_tokens": 408379063.0, "step": 10699 }, { "epoch": 1.3611499809184582, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.182183265686035, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8549028635025024, "num_tokens": 408415856.0, "step": 10700 }, { "epoch": 1.3612771911970487, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.225971221923828, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8729774951934814, "num_tokens": 408451522.0, "step": 10701 }, { "epoch": 1.3614044014756392, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.199599266052246, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.873814046382904, "num_tokens": 408494605.0, "step": 10702 }, { "epoch": 1.3615316117542298, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.20055103302002, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8668839931488037, "num_tokens": 408536373.0, "step": 10703 }, { "epoch": 1.3616588220328203, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.214560508728027, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8570797443389893, "num_tokens": 408574724.0, "step": 10704 }, { "epoch": 1.3617860323114108, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.211810111999512, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8607090711593628, "num_tokens": 408615057.0, "step": 10705 }, { "epoch": 1.3619132425900013, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.247355461120605, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8563213348388672, "num_tokens": 408651324.0, "step": 10706 }, { "epoch": 1.3620404528685919, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.190653800964355, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8740626573562622, "num_tokens": 408690533.0, "step": 10707 }, { "epoch": 1.3621676631471824, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.275418281555176, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8607879877090454, "num_tokens": 408725970.0, "step": 10708 }, { "epoch": 1.362294873425773, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.215664863586426, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8765617609024048, "num_tokens": 408766550.0, "step": 10709 }, { "epoch": 1.3624220837043632, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.229069709777832, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8726532459259033, "num_tokens": 408804320.0, "step": 10710 }, { "epoch": 1.3625492939829538, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.236763000488281, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8712824583053589, "num_tokens": 408840902.0, "step": 10711 }, { "epoch": 1.3626765042615443, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.18899154663086, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8672208786010742, "num_tokens": 408879346.0, "step": 10712 }, { "epoch": 1.3628037145401348, "ewc_loss": 0.02783203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 12.166054725646973, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8805612921714783, "num_tokens": 408919336.0, "step": 10713 }, { "epoch": 1.3629309248187254, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.233854293823242, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8629359006881714, "num_tokens": 408961547.0, "step": 10714 }, { "epoch": 1.3630581350973159, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.17989730834961, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8841416835784912, "num_tokens": 409003573.0, "step": 10715 }, { "epoch": 1.3631853453759064, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.218656539916992, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8619163036346436, "num_tokens": 409045560.0, "step": 10716 }, { "epoch": 1.363312555654497, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.187408447265625, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8733054995536804, "num_tokens": 409087245.0, "step": 10717 }, { "epoch": 1.3634397659330875, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.179281234741211, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.877048134803772, "num_tokens": 409134177.0, "step": 10718 }, { "epoch": 1.3635669762116778, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.202845573425293, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8550874590873718, "num_tokens": 409169495.0, "step": 10719 }, { "epoch": 1.3636941864902683, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.251852035522461, "learning_rate": 1e-06, "loss": 0.4814, "mean_token_accuracy": 0.8437946438789368, "num_tokens": 409207781.0, "step": 10720 }, { "epoch": 1.3638213967688588, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.211735725402832, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8692613244056702, "num_tokens": 409245411.0, "step": 10721 }, { "epoch": 1.3639486070474494, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.227527618408203, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8764157295227051, "num_tokens": 409281757.0, "step": 10722 }, { "epoch": 1.3640758173260399, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.168478012084961, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8571304678916931, "num_tokens": 409321317.0, "step": 10723 }, { "epoch": 1.3642030276046304, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.280447959899902, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8748193383216858, "num_tokens": 409358919.0, "step": 10724 }, { "epoch": 1.364330237883221, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.211485862731934, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8600542545318604, "num_tokens": 409394355.0, "step": 10725 }, { "epoch": 1.3644574481618115, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.171900749206543, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8776830434799194, "num_tokens": 409433781.0, "step": 10726 }, { "epoch": 1.364584658440402, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.185514450073242, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8587042093276978, "num_tokens": 409471728.0, "step": 10727 }, { "epoch": 1.3647118687189925, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.227120399475098, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.8863545060157776, "num_tokens": 409502294.0, "step": 10728 }, { "epoch": 1.364839078997583, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.17896556854248, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8702631592750549, "num_tokens": 409545878.0, "step": 10729 }, { "epoch": 1.3649662892761736, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.203378677368164, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8698438405990601, "num_tokens": 409584753.0, "step": 10730 }, { "epoch": 1.3650934995547641, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.261297225952148, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8729803562164307, "num_tokens": 409613725.0, "step": 10731 }, { "epoch": 1.3652207098333546, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.210367202758789, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8618161082267761, "num_tokens": 409652443.0, "step": 10732 }, { "epoch": 1.3653479201119452, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.220026016235352, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.874590277671814, "num_tokens": 409690683.0, "step": 10733 }, { "epoch": 1.3654751303905355, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.20469856262207, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8689290881156921, "num_tokens": 409727502.0, "step": 10734 }, { "epoch": 1.365602340669126, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.19397258758545, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8640636205673218, "num_tokens": 409764592.0, "step": 10735 }, { "epoch": 1.3657295509477165, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.19505500793457, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8468642830848694, "num_tokens": 409809645.0, "step": 10736 }, { "epoch": 1.365856761226307, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.179715156555176, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8685119152069092, "num_tokens": 409848048.0, "step": 10737 }, { "epoch": 1.3659839715048976, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.20378303527832, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8623899221420288, "num_tokens": 409884176.0, "step": 10738 }, { "epoch": 1.3661111817834881, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.209607124328613, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8717150688171387, "num_tokens": 409922630.0, "step": 10739 }, { "epoch": 1.3662383920620786, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.19999885559082, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.874321460723877, "num_tokens": 409961876.0, "step": 10740 }, { "epoch": 1.3663656023406692, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.19866943359375, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8766829371452332, "num_tokens": 409995432.0, "step": 10741 }, { "epoch": 1.3664928126192597, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.234415054321289, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8669412732124329, "num_tokens": 410034876.0, "step": 10742 }, { "epoch": 1.3666200228978502, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.149615287780762, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8719611167907715, "num_tokens": 410079190.0, "step": 10743 }, { "epoch": 1.3667472331764405, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.231974601745605, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8721984028816223, "num_tokens": 410112475.0, "step": 10744 }, { "epoch": 1.366874443455031, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.215290069580078, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8833850622177124, "num_tokens": 410154583.0, "step": 10745 }, { "epoch": 1.3670016537336216, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.215888977050781, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8678404092788696, "num_tokens": 410191631.0, "step": 10746 }, { "epoch": 1.3671288640122121, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.17912769317627, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8459798097610474, "num_tokens": 410232435.0, "step": 10747 }, { "epoch": 1.3672560742908026, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.219995498657227, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8725988864898682, "num_tokens": 410272940.0, "step": 10748 }, { "epoch": 1.3673832845693932, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.209410667419434, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8639229536056519, "num_tokens": 410312807.0, "step": 10749 }, { "epoch": 1.3675104948479837, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.1716947555542, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8737280368804932, "num_tokens": 410352433.0, "step": 10750 }, { "epoch": 1.3676377051265742, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.21941089630127, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8689623475074768, "num_tokens": 410394843.0, "step": 10751 }, { "epoch": 1.3677649154051648, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.148115158081055, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8708140850067139, "num_tokens": 410439353.0, "step": 10752 }, { "epoch": 1.3678921256837553, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.27232837677002, "learning_rate": 1e-06, "loss": 0.4834, "mean_token_accuracy": 0.8472892045974731, "num_tokens": 410485789.0, "step": 10753 }, { "epoch": 1.3680193359623458, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.172599792480469, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8530399799346924, "num_tokens": 410523385.0, "step": 10754 }, { "epoch": 1.3681465462409363, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.341657638549805, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8618237972259521, "num_tokens": 410553340.0, "step": 10755 }, { "epoch": 1.3682737565195269, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.154251098632812, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8764980435371399, "num_tokens": 410590041.0, "step": 10756 }, { "epoch": 1.3684009667981174, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.250737190246582, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8705250024795532, "num_tokens": 410627703.0, "step": 10757 }, { "epoch": 1.368528177076708, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.187689781188965, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8601205348968506, "num_tokens": 410662555.0, "step": 10758 }, { "epoch": 1.3686553873552982, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.200695991516113, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8503605127334595, "num_tokens": 410706026.0, "step": 10759 }, { "epoch": 1.3687825976338888, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.198108673095703, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8666642904281616, "num_tokens": 410743275.0, "step": 10760 }, { "epoch": 1.3689098079124793, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.17811107635498, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8772210478782654, "num_tokens": 410781220.0, "step": 10761 }, { "epoch": 1.3690370181910698, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.186324119567871, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8771119713783264, "num_tokens": 410820014.0, "step": 10762 }, { "epoch": 1.3691642284696603, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.215470314025879, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8632892966270447, "num_tokens": 410858721.0, "step": 10763 }, { "epoch": 1.3692914387482509, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.190221786499023, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8737349510192871, "num_tokens": 410898527.0, "step": 10764 }, { "epoch": 1.3694186490268414, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.224501609802246, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8694272637367249, "num_tokens": 410938426.0, "step": 10765 }, { "epoch": 1.369545859305432, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.24322509765625, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.873910665512085, "num_tokens": 410969940.0, "step": 10766 }, { "epoch": 1.3696730695840225, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.209456443786621, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8520563244819641, "num_tokens": 411008859.0, "step": 10767 }, { "epoch": 1.3698002798626128, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.229118347167969, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8819409608840942, "num_tokens": 411043516.0, "step": 10768 }, { "epoch": 1.3699274901412033, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.234684944152832, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8614660501480103, "num_tokens": 411079092.0, "step": 10769 }, { "epoch": 1.3700547004197938, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.137097358703613, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8726451992988586, "num_tokens": 411117360.0, "step": 10770 }, { "epoch": 1.3701819106983844, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.290313720703125, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8581453561782837, "num_tokens": 411153615.0, "step": 10771 }, { "epoch": 1.3703091209769749, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.213149070739746, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8702029585838318, "num_tokens": 411193571.0, "step": 10772 }, { "epoch": 1.3704363312555654, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.181818008422852, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8496509194374084, "num_tokens": 411230427.0, "step": 10773 }, { "epoch": 1.370563541534156, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.223873138427734, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8665724992752075, "num_tokens": 411271251.0, "step": 10774 }, { "epoch": 1.3706907518127465, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.220887184143066, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8699555993080139, "num_tokens": 411304906.0, "step": 10775 }, { "epoch": 1.370817962091337, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.20096492767334, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.885784924030304, "num_tokens": 411341417.0, "step": 10776 }, { "epoch": 1.3709451723699275, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.21389389038086, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8692585229873657, "num_tokens": 411382118.0, "step": 10777 }, { "epoch": 1.371072382648518, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.214255332946777, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8583168983459473, "num_tokens": 411426702.0, "step": 10778 }, { "epoch": 1.3711995929271086, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.262056350708008, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8672715425491333, "num_tokens": 411461689.0, "step": 10779 }, { "epoch": 1.371326803205699, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.223689079284668, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8682940602302551, "num_tokens": 411499613.0, "step": 10780 }, { "epoch": 1.3714540134842896, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.244280815124512, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.867462158203125, "num_tokens": 411538422.0, "step": 10781 }, { "epoch": 1.3715812237628802, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.235057830810547, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8812524080276489, "num_tokens": 411572401.0, "step": 10782 }, { "epoch": 1.3717084340414705, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.148221015930176, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8751596212387085, "num_tokens": 411611268.0, "step": 10783 }, { "epoch": 1.371835644320061, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.202920913696289, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8682705760002136, "num_tokens": 411648837.0, "step": 10784 }, { "epoch": 1.3719628545986515, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.189343452453613, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.871872067451477, "num_tokens": 411690327.0, "step": 10785 }, { "epoch": 1.372090064877242, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.216110229492188, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8675969243049622, "num_tokens": 411729665.0, "step": 10786 }, { "epoch": 1.3722172751558326, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.289899826049805, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8812701106071472, "num_tokens": 411764716.0, "step": 10787 }, { "epoch": 1.3723444854344231, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.186238288879395, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8697424530982971, "num_tokens": 411805560.0, "step": 10788 }, { "epoch": 1.3724716957130136, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.273197174072266, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8727245330810547, "num_tokens": 411841151.0, "step": 10789 }, { "epoch": 1.3725989059916042, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.261107444763184, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8753679990768433, "num_tokens": 411875088.0, "step": 10790 }, { "epoch": 1.3727261162701947, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.200977325439453, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8691978454589844, "num_tokens": 411914554.0, "step": 10791 }, { "epoch": 1.3728533265487852, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.263514518737793, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8692465424537659, "num_tokens": 411952237.0, "step": 10792 }, { "epoch": 1.3729805368273755, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.226387977600098, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8627306818962097, "num_tokens": 411989676.0, "step": 10793 }, { "epoch": 1.373107747105966, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.226563453674316, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8815540075302124, "num_tokens": 412025804.0, "step": 10794 }, { "epoch": 1.3732349573845566, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.224688529968262, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8845733404159546, "num_tokens": 412060085.0, "step": 10795 }, { "epoch": 1.3733621676631471, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.209285736083984, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.874630331993103, "num_tokens": 412100750.0, "step": 10796 }, { "epoch": 1.3734893779417376, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.190364837646484, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8713492155075073, "num_tokens": 412145829.0, "step": 10797 }, { "epoch": 1.3736165882203282, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.230154991149902, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8732948303222656, "num_tokens": 412187607.0, "step": 10798 }, { "epoch": 1.3737437984989187, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.258074760437012, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8558437824249268, "num_tokens": 412227214.0, "step": 10799 }, { "epoch": 1.3738710087775092, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.249037742614746, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8718165159225464, "num_tokens": 412261929.0, "step": 10800 }, { "epoch": 1.3739982190560998, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.219420433044434, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8649125099182129, "num_tokens": 412299441.0, "step": 10801 }, { "epoch": 1.3741254293346903, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.210664749145508, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8675158023834229, "num_tokens": 412339248.0, "step": 10802 }, { "epoch": 1.3742526396132808, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.235831260681152, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8672151565551758, "num_tokens": 412377224.0, "step": 10803 }, { "epoch": 1.3743798498918713, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.242679595947266, "learning_rate": 1e-06, "loss": 0.4794, "mean_token_accuracy": 0.846538782119751, "num_tokens": 412415331.0, "step": 10804 }, { "epoch": 1.3745070601704619, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.189372062683105, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8683053255081177, "num_tokens": 412455604.0, "step": 10805 }, { "epoch": 1.3746342704490524, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.280702590942383, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8756310939788818, "num_tokens": 412489295.0, "step": 10806 }, { "epoch": 1.374761480727643, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.203851699829102, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8640757203102112, "num_tokens": 412525750.0, "step": 10807 }, { "epoch": 1.3748886910062332, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.219584465026855, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.878048300743103, "num_tokens": 412567163.0, "step": 10808 }, { "epoch": 1.3750159012848238, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.2821044921875, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8753523826599121, "num_tokens": 412598789.0, "step": 10809 }, { "epoch": 1.3751431115634143, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.256942749023438, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.871881365776062, "num_tokens": 412632588.0, "step": 10810 }, { "epoch": 1.3752703218420048, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.249406814575195, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8677058219909668, "num_tokens": 412676369.0, "step": 10811 }, { "epoch": 1.3753975321205953, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.229853630065918, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8832398653030396, "num_tokens": 412712674.0, "step": 10812 }, { "epoch": 1.3755247423991859, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.247847557067871, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8631718158721924, "num_tokens": 412753531.0, "step": 10813 }, { "epoch": 1.3756519526777764, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.25466251373291, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8664622902870178, "num_tokens": 412799450.0, "step": 10814 }, { "epoch": 1.375779162956367, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.217207908630371, "learning_rate": 1e-06, "loss": 0.3407, "mean_token_accuracy": 0.8908047676086426, "num_tokens": 412836858.0, "step": 10815 }, { "epoch": 1.3759063732349575, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.295197486877441, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8714656829833984, "num_tokens": 412871169.0, "step": 10816 }, { "epoch": 1.3760335835135478, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.249372482299805, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8737257719039917, "num_tokens": 412909281.0, "step": 10817 }, { "epoch": 1.3761607937921383, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.223862648010254, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8697216510772705, "num_tokens": 412950126.0, "step": 10818 }, { "epoch": 1.3762880040707288, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.222001075744629, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8568388223648071, "num_tokens": 412988644.0, "step": 10819 }, { "epoch": 1.3764152143493193, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.199041366577148, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8551074266433716, "num_tokens": 413028403.0, "step": 10820 }, { "epoch": 1.3765424246279099, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.263378143310547, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8598254919052124, "num_tokens": 413067429.0, "step": 10821 }, { "epoch": 1.3766696349065004, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.1694917678833, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8654842376708984, "num_tokens": 413101630.0, "step": 10822 }, { "epoch": 1.376796845185091, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.270689010620117, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8625292778015137, "num_tokens": 413147757.0, "step": 10823 }, { "epoch": 1.3769240554636815, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.186614990234375, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8717395663261414, "num_tokens": 413186803.0, "step": 10824 }, { "epoch": 1.377051265742272, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.27362060546875, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8628171682357788, "num_tokens": 413227201.0, "step": 10825 }, { "epoch": 1.3771784760208625, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.180034637451172, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8668861985206604, "num_tokens": 413270594.0, "step": 10826 }, { "epoch": 1.377305686299453, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.24086856842041, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8514869213104248, "num_tokens": 413307345.0, "step": 10827 }, { "epoch": 1.3774328965780436, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.237892150878906, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8731805086135864, "num_tokens": 413344310.0, "step": 10828 }, { "epoch": 1.377560106856634, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.198800086975098, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8878830671310425, "num_tokens": 413382047.0, "step": 10829 }, { "epoch": 1.3776873171352246, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.218886375427246, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8660522699356079, "num_tokens": 413424658.0, "step": 10830 }, { "epoch": 1.3778145274138152, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.215827941894531, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8768734931945801, "num_tokens": 413458386.0, "step": 10831 }, { "epoch": 1.3779417376924055, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.249387741088867, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8560445308685303, "num_tokens": 413492530.0, "step": 10832 }, { "epoch": 1.378068947970996, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.231792449951172, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8797985315322876, "num_tokens": 413528894.0, "step": 10833 }, { "epoch": 1.3781961582495865, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.21403694152832, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8647326231002808, "num_tokens": 413564765.0, "step": 10834 }, { "epoch": 1.378323368528177, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.291885375976562, "learning_rate": 1e-06, "loss": 0.4851, "mean_token_accuracy": 0.8411292433738708, "num_tokens": 413598122.0, "step": 10835 }, { "epoch": 1.3784505788067676, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.28605842590332, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.852679967880249, "num_tokens": 413634519.0, "step": 10836 }, { "epoch": 1.378577789085358, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.21053695678711, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8607543706893921, "num_tokens": 413674021.0, "step": 10837 }, { "epoch": 1.3787049993639486, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.16032600402832, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8747310638427734, "num_tokens": 413713804.0, "step": 10838 }, { "epoch": 1.3788322096425392, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.249120712280273, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.875326931476593, "num_tokens": 413755665.0, "step": 10839 }, { "epoch": 1.3789594199211297, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.216715812683105, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8682304620742798, "num_tokens": 413790872.0, "step": 10840 }, { "epoch": 1.3790866301997202, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.210626602172852, "learning_rate": 1e-06, "loss": 0.3463, "mean_token_accuracy": 0.887700080871582, "num_tokens": 413830798.0, "step": 10841 }, { "epoch": 1.3792138404783105, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.208359718322754, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8732067346572876, "num_tokens": 413865960.0, "step": 10842 }, { "epoch": 1.379341050756901, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.286813735961914, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8791332244873047, "num_tokens": 413898222.0, "step": 10843 }, { "epoch": 1.3794682610354916, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.266759872436523, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8633646965026855, "num_tokens": 413931839.0, "step": 10844 }, { "epoch": 1.3795954713140821, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.21019172668457, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8600337505340576, "num_tokens": 413963815.0, "step": 10845 }, { "epoch": 1.3797226815926726, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.257511138916016, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8768771886825562, "num_tokens": 414002476.0, "step": 10846 }, { "epoch": 1.3798498918712632, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.224344253540039, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8634655475616455, "num_tokens": 414041121.0, "step": 10847 }, { "epoch": 1.3799771021498537, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.285080909729004, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8563158512115479, "num_tokens": 414080447.0, "step": 10848 }, { "epoch": 1.3801043124284442, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.220076560974121, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8586355447769165, "num_tokens": 414108925.0, "step": 10849 }, { "epoch": 1.3802315227070348, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.281510353088379, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8735843300819397, "num_tokens": 414154296.0, "step": 10850 }, { "epoch": 1.3803587329856253, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.20557975769043, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8566504716873169, "num_tokens": 414194612.0, "step": 10851 }, { "epoch": 1.3804859432642158, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.247499465942383, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8716230392456055, "num_tokens": 414229939.0, "step": 10852 }, { "epoch": 1.3806131535428063, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.292813301086426, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.865352213382721, "num_tokens": 414265193.0, "step": 10853 }, { "epoch": 1.3807403638213969, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.205077171325684, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8664798736572266, "num_tokens": 414301471.0, "step": 10854 }, { "epoch": 1.3808675740999874, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.260770797729492, "learning_rate": 1e-06, "loss": 0.4948, "mean_token_accuracy": 0.839789628982544, "num_tokens": 414338755.0, "step": 10855 }, { "epoch": 1.380994784378578, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.25475788116455, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8697385787963867, "num_tokens": 414379996.0, "step": 10856 }, { "epoch": 1.3811219946571682, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.260026931762695, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8713173866271973, "num_tokens": 414421905.0, "step": 10857 }, { "epoch": 1.3812492049357588, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.28199577331543, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8741458058357239, "num_tokens": 414465149.0, "step": 10858 }, { "epoch": 1.3813764152143493, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.245323181152344, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8779978156089783, "num_tokens": 414504086.0, "step": 10859 }, { "epoch": 1.3815036254929398, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.23794937133789, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8526327610015869, "num_tokens": 414545392.0, "step": 10860 }, { "epoch": 1.3816308357715303, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.302448272705078, "learning_rate": 1e-06, "loss": 0.5144, "mean_token_accuracy": 0.8339014053344727, "num_tokens": 414586445.0, "step": 10861 }, { "epoch": 1.3817580460501209, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.241312026977539, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8771193027496338, "num_tokens": 414618467.0, "step": 10862 }, { "epoch": 1.3818852563287114, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.26266098022461, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8787391781806946, "num_tokens": 414651253.0, "step": 10863 }, { "epoch": 1.382012466607302, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.238655090332031, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8679199814796448, "num_tokens": 414692016.0, "step": 10864 }, { "epoch": 1.3821396768858925, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.253748893737793, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8736947178840637, "num_tokens": 414731840.0, "step": 10865 }, { "epoch": 1.3822668871644828, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.266284942626953, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.869522213935852, "num_tokens": 414767110.0, "step": 10866 }, { "epoch": 1.3823940974430733, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.260175704956055, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8741950392723083, "num_tokens": 414799340.0, "step": 10867 }, { "epoch": 1.3825213077216638, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.280435562133789, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.871204137802124, "num_tokens": 414840915.0, "step": 10868 }, { "epoch": 1.3826485180002543, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.29675006866455, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8565654754638672, "num_tokens": 414877298.0, "step": 10869 }, { "epoch": 1.3827757282788449, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.251035690307617, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8758684396743774, "num_tokens": 414916431.0, "step": 10870 }, { "epoch": 1.3829029385574354, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.278572082519531, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8668875098228455, "num_tokens": 414958148.0, "step": 10871 }, { "epoch": 1.383030148836026, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.223864555358887, "learning_rate": 1e-06, "loss": 0.3353, "mean_token_accuracy": 0.8932657241821289, "num_tokens": 414997008.0, "step": 10872 }, { "epoch": 1.3831573591146165, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.263107299804688, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8550742268562317, "num_tokens": 415042893.0, "step": 10873 }, { "epoch": 1.383284569393207, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.285858154296875, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8663440942764282, "num_tokens": 415077476.0, "step": 10874 }, { "epoch": 1.3834117796717975, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.2579927444458, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8639176487922668, "num_tokens": 415124329.0, "step": 10875 }, { "epoch": 1.383538989950388, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.237338066101074, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8794847130775452, "num_tokens": 415164101.0, "step": 10876 }, { "epoch": 1.3836662002289786, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.257424354553223, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.869350790977478, "num_tokens": 415203792.0, "step": 10877 }, { "epoch": 1.383793410507569, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.335342407226562, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8657885789871216, "num_tokens": 415233754.0, "step": 10878 }, { "epoch": 1.3839206207861596, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.252959251403809, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8753365874290466, "num_tokens": 415269567.0, "step": 10879 }, { "epoch": 1.3840478310647502, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.212692260742188, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8730161786079407, "num_tokens": 415312660.0, "step": 10880 }, { "epoch": 1.3841750413433405, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.203064918518066, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8780068755149841, "num_tokens": 415355377.0, "step": 10881 }, { "epoch": 1.384302251621931, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.237411499023438, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8801474571228027, "num_tokens": 415393109.0, "step": 10882 }, { "epoch": 1.3844294619005215, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.227743148803711, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8799703121185303, "num_tokens": 415428650.0, "step": 10883 }, { "epoch": 1.384556672179112, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.295489311218262, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8662530183792114, "num_tokens": 415467166.0, "step": 10884 }, { "epoch": 1.3846838824577026, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.21605396270752, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8815576434135437, "num_tokens": 415505633.0, "step": 10885 }, { "epoch": 1.384811092736293, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.313603401184082, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8681588172912598, "num_tokens": 415546254.0, "step": 10886 }, { "epoch": 1.3849383030148836, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.275989532470703, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8677192330360413, "num_tokens": 415584999.0, "step": 10887 }, { "epoch": 1.3850655132934742, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.271258354187012, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8717507719993591, "num_tokens": 415624608.0, "step": 10888 }, { "epoch": 1.3851927235720647, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.224924087524414, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8558663725852966, "num_tokens": 415663405.0, "step": 10889 }, { "epoch": 1.385319933850655, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.246633529663086, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8828543424606323, "num_tokens": 415698934.0, "step": 10890 }, { "epoch": 1.3854471441292455, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.301997184753418, "learning_rate": 1e-06, "loss": 0.4976, "mean_token_accuracy": 0.8412856459617615, "num_tokens": 415736973.0, "step": 10891 }, { "epoch": 1.385574354407836, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.236746788024902, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.87250816822052, "num_tokens": 415778364.0, "step": 10892 }, { "epoch": 1.3857015646864266, "ewc_loss": 0.028076171875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 12.27904224395752, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8848913908004761, "num_tokens": 415817502.0, "step": 10893 }, { "epoch": 1.385828774965017, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.214418411254883, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.864824652671814, "num_tokens": 415860364.0, "step": 10894 }, { "epoch": 1.3859559852436076, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.283769607543945, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8724350333213806, "num_tokens": 415896816.0, "step": 10895 }, { "epoch": 1.3860831955221982, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.260499954223633, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8522339463233948, "num_tokens": 415935173.0, "step": 10896 }, { "epoch": 1.3862104058007887, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.24873161315918, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8783924579620361, "num_tokens": 415977043.0, "step": 10897 }, { "epoch": 1.3863376160793792, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.254379272460938, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8783129453659058, "num_tokens": 416014376.0, "step": 10898 }, { "epoch": 1.3864648263579697, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.266353607177734, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8551832437515259, "num_tokens": 416055648.0, "step": 10899 }, { "epoch": 1.3865920366365603, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.2843599319458, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8611767292022705, "num_tokens": 416087062.0, "step": 10900 }, { "epoch": 1.3867192469151508, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.307418823242188, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8758483529090881, "num_tokens": 416122559.0, "step": 10901 }, { "epoch": 1.3868464571937413, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.319421768188477, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8619312047958374, "num_tokens": 416155228.0, "step": 10902 }, { "epoch": 1.3869736674723319, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.248223304748535, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.854469895362854, "num_tokens": 416191190.0, "step": 10903 }, { "epoch": 1.3871008777509224, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.300448417663574, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.872604489326477, "num_tokens": 416228009.0, "step": 10904 }, { "epoch": 1.387228088029513, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.262636184692383, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8635738492012024, "num_tokens": 416263657.0, "step": 10905 }, { "epoch": 1.3873552983081032, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.281041145324707, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.86210036277771, "num_tokens": 416305187.0, "step": 10906 }, { "epoch": 1.3874825085866938, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.286748886108398, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8662401437759399, "num_tokens": 416342578.0, "step": 10907 }, { "epoch": 1.3876097188652843, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.24870777130127, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8664520382881165, "num_tokens": 416378969.0, "step": 10908 }, { "epoch": 1.3877369291438748, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.26831340789795, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8666525483131409, "num_tokens": 416412958.0, "step": 10909 }, { "epoch": 1.3878641394224653, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.250309944152832, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8499532341957092, "num_tokens": 416445643.0, "step": 10910 }, { "epoch": 1.3879913497010559, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.296477317810059, "learning_rate": 1e-06, "loss": 0.4498, "mean_token_accuracy": 0.8548741340637207, "num_tokens": 416484891.0, "step": 10911 }, { "epoch": 1.3881185599796464, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.290437698364258, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8550408482551575, "num_tokens": 416520504.0, "step": 10912 }, { "epoch": 1.388245770258237, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.198058128356934, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8690586090087891, "num_tokens": 416561877.0, "step": 10913 }, { "epoch": 1.3883729805368275, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.277572631835938, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8700658679008484, "num_tokens": 416597676.0, "step": 10914 }, { "epoch": 1.3885001908154178, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.303603172302246, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8746640086174011, "num_tokens": 416630557.0, "step": 10915 }, { "epoch": 1.3886274010940083, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.285360336303711, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8716834783554077, "num_tokens": 416668604.0, "step": 10916 }, { "epoch": 1.3887546113725988, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.247648239135742, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8482219576835632, "num_tokens": 416710246.0, "step": 10917 }, { "epoch": 1.3888818216511893, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.266120910644531, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8686343431472778, "num_tokens": 416744962.0, "step": 10918 }, { "epoch": 1.3890090319297799, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.33870792388916, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.869703471660614, "num_tokens": 416779483.0, "step": 10919 }, { "epoch": 1.3891362422083704, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.201793670654297, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.880994975566864, "num_tokens": 416822135.0, "step": 10920 }, { "epoch": 1.389263452486961, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.279129981994629, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8662636280059814, "num_tokens": 416860383.0, "step": 10921 }, { "epoch": 1.3893906627655515, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.297202110290527, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8644694685935974, "num_tokens": 416896960.0, "step": 10922 }, { "epoch": 1.389517873044142, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.314496040344238, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8796069025993347, "num_tokens": 416931981.0, "step": 10923 }, { "epoch": 1.3896450833227325, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.225110054016113, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8775621652603149, "num_tokens": 416968179.0, "step": 10924 }, { "epoch": 1.389772293601323, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.332377433776855, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8729575276374817, "num_tokens": 417008879.0, "step": 10925 }, { "epoch": 1.3898995038799136, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.254484176635742, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.878919243812561, "num_tokens": 417042497.0, "step": 10926 }, { "epoch": 1.390026714158504, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.303911209106445, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8694559931755066, "num_tokens": 417084396.0, "step": 10927 }, { "epoch": 1.3901539244370946, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.247810363769531, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8817826509475708, "num_tokens": 417121014.0, "step": 10928 }, { "epoch": 1.3902811347156852, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.294997215270996, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8676362037658691, "num_tokens": 417160169.0, "step": 10929 }, { "epoch": 1.3904083449942755, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.33450698852539, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8526144623756409, "num_tokens": 417195024.0, "step": 10930 }, { "epoch": 1.390535555272866, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.230566024780273, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8740358352661133, "num_tokens": 417231344.0, "step": 10931 }, { "epoch": 1.3906627655514565, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.311970710754395, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8604038953781128, "num_tokens": 417270266.0, "step": 10932 }, { "epoch": 1.390789975830047, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.314879417419434, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8666810989379883, "num_tokens": 417308292.0, "step": 10933 }, { "epoch": 1.3909171861086376, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.24687671661377, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8665623664855957, "num_tokens": 417346358.0, "step": 10934 }, { "epoch": 1.391044396387228, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.31181526184082, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8689221143722534, "num_tokens": 417381924.0, "step": 10935 }, { "epoch": 1.3911716066658186, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.26191520690918, "learning_rate": 1e-06, "loss": 0.4866, "mean_token_accuracy": 0.8423879146575928, "num_tokens": 417417779.0, "step": 10936 }, { "epoch": 1.3912988169444092, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.232393264770508, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8612270355224609, "num_tokens": 417459076.0, "step": 10937 }, { "epoch": 1.3914260272229997, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.232646942138672, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8682516813278198, "num_tokens": 417499422.0, "step": 10938 }, { "epoch": 1.39155323750159, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.271379470825195, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8689266443252563, "num_tokens": 417541776.0, "step": 10939 }, { "epoch": 1.3916804477801805, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.325228691101074, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8564209938049316, "num_tokens": 417576469.0, "step": 10940 }, { "epoch": 1.391807658058771, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.244880676269531, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8687422275543213, "num_tokens": 417611726.0, "step": 10941 }, { "epoch": 1.3919348683373616, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.307252883911133, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8604336977005005, "num_tokens": 417650135.0, "step": 10942 }, { "epoch": 1.392062078615952, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.33462905883789, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8498126864433289, "num_tokens": 417690981.0, "step": 10943 }, { "epoch": 1.3921892888945426, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.286590576171875, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8560796976089478, "num_tokens": 417732152.0, "step": 10944 }, { "epoch": 1.3923164991731332, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.252638816833496, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8574628829956055, "num_tokens": 417771814.0, "step": 10945 }, { "epoch": 1.3924437094517237, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.267114639282227, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.858078122138977, "num_tokens": 417810663.0, "step": 10946 }, { "epoch": 1.3925709197303142, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.24591064453125, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8718558549880981, "num_tokens": 417845664.0, "step": 10947 }, { "epoch": 1.3926981300089047, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.282661437988281, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8599466681480408, "num_tokens": 417889617.0, "step": 10948 }, { "epoch": 1.3928253402874953, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.25568675994873, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8658080697059631, "num_tokens": 417930820.0, "step": 10949 }, { "epoch": 1.3929525505660858, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.259732246398926, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8694599866867065, "num_tokens": 417966799.0, "step": 10950 }, { "epoch": 1.3930797608446763, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.289868354797363, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8610770106315613, "num_tokens": 418009328.0, "step": 10951 }, { "epoch": 1.3932069711232669, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.275971412658691, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8699514865875244, "num_tokens": 418048548.0, "step": 10952 }, { "epoch": 1.3933341814018574, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.211163520812988, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8655643463134766, "num_tokens": 418094947.0, "step": 10953 }, { "epoch": 1.393461391680448, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.302919387817383, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8585402965545654, "num_tokens": 418137144.0, "step": 10954 }, { "epoch": 1.3935886019590382, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.225133895874023, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8895334601402283, "num_tokens": 418171449.0, "step": 10955 }, { "epoch": 1.3937158122376287, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.294379234313965, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8702871799468994, "num_tokens": 418209854.0, "step": 10956 }, { "epoch": 1.3938430225162193, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.279854774475098, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8754035234451294, "num_tokens": 418245121.0, "step": 10957 }, { "epoch": 1.3939702327948098, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.281864166259766, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8716000318527222, "num_tokens": 418284158.0, "step": 10958 }, { "epoch": 1.3940974430734003, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.242852210998535, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.874555766582489, "num_tokens": 418324091.0, "step": 10959 }, { "epoch": 1.3942246533519909, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.249269485473633, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8678308129310608, "num_tokens": 418364764.0, "step": 10960 }, { "epoch": 1.3943518636305814, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.222579956054688, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8655675649642944, "num_tokens": 418403974.0, "step": 10961 }, { "epoch": 1.394479073909172, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.245469093322754, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8580203652381897, "num_tokens": 418446341.0, "step": 10962 }, { "epoch": 1.3946062841877624, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.263707160949707, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8678876161575317, "num_tokens": 418490989.0, "step": 10963 }, { "epoch": 1.3947334944663528, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.280811309814453, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8618798851966858, "num_tokens": 418524633.0, "step": 10964 }, { "epoch": 1.3948607047449433, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.244744300842285, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8669015169143677, "num_tokens": 418562576.0, "step": 10965 }, { "epoch": 1.3949879150235338, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.24864387512207, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8668336868286133, "num_tokens": 418596537.0, "step": 10966 }, { "epoch": 1.3951151253021243, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.317338943481445, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8660910129547119, "num_tokens": 418625800.0, "step": 10967 }, { "epoch": 1.3952423355807149, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.2744140625, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.8542434573173523, "num_tokens": 418663818.0, "step": 10968 }, { "epoch": 1.3953695458593054, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.239311218261719, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8730306625366211, "num_tokens": 418700607.0, "step": 10969 }, { "epoch": 1.395496756137896, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.271350860595703, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8620598912239075, "num_tokens": 418739357.0, "step": 10970 }, { "epoch": 1.3956239664164865, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.27511978149414, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8739629983901978, "num_tokens": 418769752.0, "step": 10971 }, { "epoch": 1.395751176695077, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.255674362182617, "learning_rate": 1e-06, "loss": 0.4693, "mean_token_accuracy": 0.8488550186157227, "num_tokens": 418808196.0, "step": 10972 }, { "epoch": 1.3958783869736675, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.25475025177002, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8833212852478027, "num_tokens": 418843335.0, "step": 10973 }, { "epoch": 1.396005597252258, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.256550788879395, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8760522603988647, "num_tokens": 418876591.0, "step": 10974 }, { "epoch": 1.3961328075308486, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.29894733428955, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8554686307907104, "num_tokens": 418910482.0, "step": 10975 }, { "epoch": 1.396260017809439, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.25755500793457, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8593244552612305, "num_tokens": 418942248.0, "step": 10976 }, { "epoch": 1.3963872280880296, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.250019073486328, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.871641993522644, "num_tokens": 418977770.0, "step": 10977 }, { "epoch": 1.3965144383666201, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.270211219787598, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8609809875488281, "num_tokens": 419016464.0, "step": 10978 }, { "epoch": 1.3966416486452105, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.236546516418457, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8592244386672974, "num_tokens": 419053254.0, "step": 10979 }, { "epoch": 1.396768858923801, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.259042739868164, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8606541156768799, "num_tokens": 419088940.0, "step": 10980 }, { "epoch": 1.3968960692023915, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.253305435180664, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8666831254959106, "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": 12.199296951293945, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8699554204940796, "num_tokens": 419170921.0, "step": 10982 }, { "epoch": 1.3971504897595726, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.310510635375977, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8738889694213867, "num_tokens": 419206032.0, "step": 10983 }, { "epoch": 1.397277700038163, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.263161659240723, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8822025060653687, "num_tokens": 419244240.0, "step": 10984 }, { "epoch": 1.3974049103167536, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.31698226928711, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8759046792984009, "num_tokens": 419274902.0, "step": 10985 }, { "epoch": 1.3975321205953442, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.228913307189941, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8798710107803345, "num_tokens": 419312295.0, "step": 10986 }, { "epoch": 1.3976593308739347, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.296919822692871, "learning_rate": 1e-06, "loss": 0.4651, "mean_token_accuracy": 0.8502969741821289, "num_tokens": 419350999.0, "step": 10987 }, { "epoch": 1.397786541152525, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.303966522216797, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8685088157653809, "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": 12.182723045349121, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8745613694190979, "num_tokens": 419430353.0, "step": 10989 }, { "epoch": 1.398040961709706, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.31884765625, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8636683225631714, "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": 12.238568305969238, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8772146701812744, "num_tokens": 419498727.0, "step": 10991 }, { "epoch": 1.398295382266887, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.306990623474121, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8592637181282043, "num_tokens": 419534496.0, "step": 10992 }, { "epoch": 1.3984225925454776, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.274490356445312, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.863974928855896, "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": 12.317423820495605, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8751990795135498, "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": 12.269956588745117, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8747972249984741, "num_tokens": 419650425.0, "step": 10995 }, { "epoch": 1.3988042233812492, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.308690071105957, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8753241300582886, "num_tokens": 419685501.0, "step": 10996 }, { "epoch": 1.3989314336598397, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.284669876098633, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8703255653381348, "num_tokens": 419724471.0, "step": 10997 }, { "epoch": 1.3990586439384303, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.250149726867676, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8756438493728638, "num_tokens": 419763683.0, "step": 10998 }, { "epoch": 1.3991858542170208, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.310203552246094, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8711267113685608, "num_tokens": 419801238.0, "step": 10999 }, { "epoch": 1.3993130644956113, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.276385307312012, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8643941283226013, "num_tokens": 419832787.0, "step": 11000 }, { "epoch": 1.3994402747742019, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.350963592529297, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8580144643783569, "num_tokens": 419864294.0, "step": 11001 }, { "epoch": 1.3995674850527924, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.300585746765137, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8705499172210693, "num_tokens": 419900493.0, "step": 11002 }, { "epoch": 1.399694695331383, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.241837501525879, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8863427639007568, "num_tokens": 419938917.0, "step": 11003 }, { "epoch": 1.3998219056099732, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.282730102539062, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8832323551177979, "num_tokens": 419978771.0, "step": 11004 }, { "epoch": 1.3999491158885637, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.32528305053711, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8665187358856201, "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": 12.257875442504883, "learning_rate": 1e-06, "loss": 0.4828, "mean_token_accuracy": 0.8472390174865723, "num_tokens": 420051618.0, "step": 11006 }, { "epoch": 1.4002035364457448, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.320551872253418, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8549762964248657, "num_tokens": 420089119.0, "step": 11007 }, { "epoch": 1.4003307467243353, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.33178424835205, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8799054026603699, "num_tokens": 420123706.0, "step": 11008 }, { "epoch": 1.4004579570029259, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.296647071838379, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8650429844856262, "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": 12.277976036071777, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8676445484161377, "num_tokens": 420198519.0, "step": 11010 }, { "epoch": 1.400712377560107, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.27548599243164, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8807665705680847, "num_tokens": 420231722.0, "step": 11011 }, { "epoch": 1.4008395878386974, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.27401351928711, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.870974063873291, "num_tokens": 420272059.0, "step": 11012 }, { "epoch": 1.4009667981172877, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.315322875976562, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8750101327896118, "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": 12.256958961486816, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8774495124816895, "num_tokens": 420345989.0, "step": 11014 }, { "epoch": 1.4012212186744688, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.263819694519043, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8746079206466675, "num_tokens": 420387194.0, "step": 11015 }, { "epoch": 1.4013484289530593, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.286734580993652, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8835813403129578, "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": 12.259540557861328, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8741089105606079, "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": 12.280821800231934, "learning_rate": 1e-06, "loss": 0.4513, "mean_token_accuracy": 0.8499882221221924, "num_tokens": 420503577.0, "step": 11018 }, { "epoch": 1.401730059788831, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.273090362548828, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8790965676307678, "num_tokens": 420538852.0, "step": 11019 }, { "epoch": 1.4018572700674214, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.272893905639648, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8657071590423584, "num_tokens": 420577495.0, "step": 11020 }, { "epoch": 1.401984480346012, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.238590240478516, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8686224222183228, "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": 12.242576599121094, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8781540393829346, "num_tokens": 420653904.0, "step": 11022 }, { "epoch": 1.402238900903193, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.21977710723877, "learning_rate": 1e-06, "loss": 0.3461, "mean_token_accuracy": 0.8919715881347656, "num_tokens": 420693539.0, "step": 11023 }, { "epoch": 1.4023661111817836, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.33405876159668, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.872070848941803, "num_tokens": 420729046.0, "step": 11024 }, { "epoch": 1.402493321460374, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.287532806396484, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8721283674240112, "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": 12.231839179992676, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8657179474830627, "num_tokens": 420806367.0, "step": 11026 }, { "epoch": 1.4027477420175551, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.283068656921387, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.875490665435791, "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": 12.21855640411377, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.873818039894104, "num_tokens": 420885833.0, "step": 11028 }, { "epoch": 1.403002162574736, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.325084686279297, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8742936849594116, "num_tokens": 420916313.0, "step": 11029 }, { "epoch": 1.4031293728533265, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.28537654876709, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8510807156562805, "num_tokens": 420957011.0, "step": 11030 }, { "epoch": 1.403256583131917, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.226340293884277, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8726915121078491, "num_tokens": 420994815.0, "step": 11031 }, { "epoch": 1.4033837934105076, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.319559097290039, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8759486675262451, "num_tokens": 421033977.0, "step": 11032 }, { "epoch": 1.403511003689098, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.289572715759277, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8728369474411011, "num_tokens": 421068871.0, "step": 11033 }, { "epoch": 1.4036382139676886, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.297676086425781, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8697917461395264, "num_tokens": 421105680.0, "step": 11034 }, { "epoch": 1.4037654242462791, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.287599563598633, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8810933828353882, "num_tokens": 421149519.0, "step": 11035 }, { "epoch": 1.4038926345248697, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.277765274047852, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.870391845703125, "num_tokens": 421190195.0, "step": 11036 }, { "epoch": 1.40401984480346, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.298746109008789, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8645932674407959, "num_tokens": 421218637.0, "step": 11037 }, { "epoch": 1.4041470550820505, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.349326133728027, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8734170198440552, "num_tokens": 421261300.0, "step": 11038 }, { "epoch": 1.404274265360641, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.238567352294922, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8659101128578186, "num_tokens": 421297250.0, "step": 11039 }, { "epoch": 1.4044014756392316, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.34498405456543, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8620618581771851, "num_tokens": 421334213.0, "step": 11040 }, { "epoch": 1.404528685917822, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.250389099121094, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8760218024253845, "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": 12.30770492553711, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8574441075325012, "num_tokens": 421408278.0, "step": 11042 }, { "epoch": 1.4047831064750032, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.260750770568848, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8611304759979248, "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": 12.378491401672363, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.868101954460144, "num_tokens": 421474014.0, "step": 11044 }, { "epoch": 1.4050375270321842, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.277385711669922, "learning_rate": 1e-06, "loss": 0.4829, "mean_token_accuracy": 0.8444656729698181, "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": 12.290749549865723, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8702899813652039, "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": 12.302050590515137, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8683056831359863, "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": 12.27723217010498, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8710588216781616, "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": 12.28860855102539, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8787913918495178, "num_tokens": 421659451.0, "step": 11049 }, { "epoch": 1.4056735784251368, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.306320190429688, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8666355609893799, "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": 12.278523445129395, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8649088740348816, "num_tokens": 421748953.0, "step": 11051 }, { "epoch": 1.405927998982318, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.32927131652832, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.851879894733429, "num_tokens": 421786614.0, "step": 11052 }, { "epoch": 1.4060552092609082, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.30608081817627, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8631129264831543, "num_tokens": 421817291.0, "step": 11053 }, { "epoch": 1.4061824195394987, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.313519477844238, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8658708333969116, "num_tokens": 421858200.0, "step": 11054 }, { "epoch": 1.4063096298180893, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.275945663452148, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8687427043914795, "num_tokens": 421903343.0, "step": 11055 }, { "epoch": 1.4064368400966798, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.285062789916992, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.884933590888977, "num_tokens": 421941426.0, "step": 11056 }, { "epoch": 1.4065640503752703, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.303257942199707, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8659769296646118, "num_tokens": 421983992.0, "step": 11057 }, { "epoch": 1.4066912606538609, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.310269355773926, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8777321577072144, "num_tokens": 422015066.0, "step": 11058 }, { "epoch": 1.4068184709324514, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.325844764709473, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8708810806274414, "num_tokens": 422052392.0, "step": 11059 }, { "epoch": 1.406945681211042, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.334942817687988, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8640334606170654, "num_tokens": 422091453.0, "step": 11060 }, { "epoch": 1.4070728914896324, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.364069938659668, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.852091908454895, "num_tokens": 422131019.0, "step": 11061 }, { "epoch": 1.4072001017682227, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.270768165588379, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8621645569801331, "num_tokens": 422172139.0, "step": 11062 }, { "epoch": 1.4073273120468133, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.349425315856934, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8754500150680542, "num_tokens": 422209708.0, "step": 11063 }, { "epoch": 1.4074545223254038, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.301498413085938, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8697789907455444, "num_tokens": 422245091.0, "step": 11064 }, { "epoch": 1.4075817326039943, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.315568923950195, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8689016103744507, "num_tokens": 422280054.0, "step": 11065 }, { "epoch": 1.4077089428825849, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.308404922485352, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.874844491481781, "num_tokens": 422310036.0, "step": 11066 }, { "epoch": 1.4078361531611754, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.251729965209961, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8533182144165039, "num_tokens": 422347914.0, "step": 11067 }, { "epoch": 1.407963363439766, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.322145462036133, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8702036738395691, "num_tokens": 422383207.0, "step": 11068 }, { "epoch": 1.4080905737183564, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.338207244873047, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8720202445983887, "num_tokens": 422414675.0, "step": 11069 }, { "epoch": 1.408217783996947, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.29553508758545, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8575763702392578, "num_tokens": 422449746.0, "step": 11070 }, { "epoch": 1.4083449942755375, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.254194259643555, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8623544573783875, "num_tokens": 422487789.0, "step": 11071 }, { "epoch": 1.408472204554128, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.279784202575684, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8744393587112427, "num_tokens": 422521750.0, "step": 11072 }, { "epoch": 1.4085994148327186, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.298361778259277, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8781486749649048, "num_tokens": 422562923.0, "step": 11073 }, { "epoch": 1.408726625111309, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.236303329467773, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8587545156478882, "num_tokens": 422606832.0, "step": 11074 }, { "epoch": 1.4088538353898996, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.364986419677734, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8725271821022034, "num_tokens": 422644465.0, "step": 11075 }, { "epoch": 1.4089810456684901, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.28054141998291, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8771598935127258, "num_tokens": 422681204.0, "step": 11076 }, { "epoch": 1.4091082559470804, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.331522941589355, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8585806488990784, "num_tokens": 422716325.0, "step": 11077 }, { "epoch": 1.409235466225671, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.282417297363281, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8579331636428833, "num_tokens": 422753644.0, "step": 11078 }, { "epoch": 1.4093626765042615, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.25609302520752, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8687942624092102, "num_tokens": 422790893.0, "step": 11079 }, { "epoch": 1.409489886782852, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.330780982971191, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8610192537307739, "num_tokens": 422829400.0, "step": 11080 }, { "epoch": 1.4096170970614426, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.246045112609863, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8775211572647095, "num_tokens": 422867386.0, "step": 11081 }, { "epoch": 1.409744307340033, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.300753593444824, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8712983131408691, "num_tokens": 422910736.0, "step": 11082 }, { "epoch": 1.4098715176186236, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.322545051574707, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8732021450996399, "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": 12.33073616027832, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8672225475311279, "num_tokens": 422985892.0, "step": 11084 }, { "epoch": 1.4101259381758047, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.350842475891113, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8613908886909485, "num_tokens": 423020075.0, "step": 11085 }, { "epoch": 1.410253148454395, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.288633346557617, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8714682459831238, "num_tokens": 423058260.0, "step": 11086 }, { "epoch": 1.4103803587329855, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.322205543518066, "learning_rate": 1e-06, "loss": 0.3354, "mean_token_accuracy": 0.890877366065979, "num_tokens": 423092158.0, "step": 11087 }, { "epoch": 1.410507569011576, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.334793090820312, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8665082454681396, "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": 12.31201457977295, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8646674156188965, "num_tokens": 423167111.0, "step": 11089 }, { "epoch": 1.410761989568757, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.345823287963867, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8607812523841858, "num_tokens": 423202060.0, "step": 11090 }, { "epoch": 1.4108891998473476, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.35159683227539, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8601694107055664, "num_tokens": 423241200.0, "step": 11091 }, { "epoch": 1.4110164101259381, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.279989242553711, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.877069354057312, "num_tokens": 423284713.0, "step": 11092 }, { "epoch": 1.4111436204045287, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.332918167114258, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.871069073677063, "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": 12.319517135620117, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8694847822189331, "num_tokens": 423360551.0, "step": 11094 }, { "epoch": 1.4113980409617097, "ewc_loss": 0.028564453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.308915138244629, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8787379264831543, "num_tokens": 423395227.0, "step": 11095 }, { "epoch": 1.4115252512403003, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.363844871520996, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8545514941215515, "num_tokens": 423434757.0, "step": 11096 }, { "epoch": 1.4116524615188908, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.370671272277832, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8857028484344482, "num_tokens": 423464599.0, "step": 11097 }, { "epoch": 1.4117796717974813, "ewc_loss": 0.0283203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.33713150024414, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8619815111160278, "num_tokens": 423504327.0, "step": 11098 }, { "epoch": 1.4119068820760718, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.335102081298828, "learning_rate": 1e-06, "loss": 0.4755, "mean_token_accuracy": 0.8495985865592957, "num_tokens": 423543114.0, "step": 11099 }, { "epoch": 1.4120340923546624, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.288086891174316, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8570088744163513, "num_tokens": 423581727.0, "step": 11100 }, { "epoch": 1.412161302633253, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.292991638183594, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8725477457046509, "num_tokens": 423617073.0, "step": 11101 }, { "epoch": 1.4122885129118432, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.296433448791504, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8726787567138672, "num_tokens": 423664922.0, "step": 11102 }, { "epoch": 1.4124157231904337, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.3348388671875, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8592965006828308, "num_tokens": 423706990.0, "step": 11103 }, { "epoch": 1.4125429334690243, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.312829971313477, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.8522595763206482, "num_tokens": 423745945.0, "step": 11104 }, { "epoch": 1.4126701437476148, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.28615951538086, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8740367889404297, "num_tokens": 423780824.0, "step": 11105 }, { "epoch": 1.4127973540262053, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.341998100280762, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8817999958992004, "num_tokens": 423818640.0, "step": 11106 }, { "epoch": 1.4129245643047958, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.328882217407227, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8776358962059021, "num_tokens": 423856657.0, "step": 11107 }, { "epoch": 1.4130517745833864, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.286340713500977, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8801586031913757, "num_tokens": 423900288.0, "step": 11108 }, { "epoch": 1.413178984861977, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.348735809326172, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8760101795196533, "num_tokens": 423935929.0, "step": 11109 }, { "epoch": 1.4133061951405674, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.30923080444336, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.867068886756897, "num_tokens": 423970719.0, "step": 11110 }, { "epoch": 1.4134334054191577, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.338037490844727, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8694419860839844, "num_tokens": 424012338.0, "step": 11111 }, { "epoch": 1.4135606156977483, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.294602394104004, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8522628545761108, "num_tokens": 424050112.0, "step": 11112 }, { "epoch": 1.4136878259763388, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.360337257385254, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8695703148841858, "num_tokens": 424092688.0, "step": 11113 }, { "epoch": 1.4138150362549293, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.268074035644531, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8671610355377197, "num_tokens": 424131355.0, "step": 11114 }, { "epoch": 1.4139422465335199, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.305888175964355, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8589603900909424, "num_tokens": 424177768.0, "step": 11115 }, { "epoch": 1.4140694568121104, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.278486251831055, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8814759254455566, "num_tokens": 424215834.0, "step": 11116 }, { "epoch": 1.414196667090701, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.299781799316406, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8725026845932007, "num_tokens": 424260002.0, "step": 11117 }, { "epoch": 1.4143238773692914, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.266960144042969, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.867012619972229, "num_tokens": 424299343.0, "step": 11118 }, { "epoch": 1.414451087647882, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.36937141418457, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8630510568618774, "num_tokens": 424340312.0, "step": 11119 }, { "epoch": 1.4145782979264725, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.269506454467773, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8771785497665405, "num_tokens": 424376658.0, "step": 11120 }, { "epoch": 1.414705508205063, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.332869529724121, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8796424865722656, "num_tokens": 424415499.0, "step": 11121 }, { "epoch": 1.4148327184836536, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.329619407653809, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8701142072677612, "num_tokens": 424453401.0, "step": 11122 }, { "epoch": 1.414959928762244, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.36249828338623, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8725744485855103, "num_tokens": 424490654.0, "step": 11123 }, { "epoch": 1.4150871390408346, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.354496955871582, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8785749673843384, "num_tokens": 424534469.0, "step": 11124 }, { "epoch": 1.4152143493194251, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.338628768920898, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8603439331054688, "num_tokens": 424570187.0, "step": 11125 }, { "epoch": 1.4153415595980154, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.29192066192627, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8493001461029053, "num_tokens": 424607898.0, "step": 11126 }, { "epoch": 1.415468769876606, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.284295082092285, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8711973428726196, "num_tokens": 424652156.0, "step": 11127 }, { "epoch": 1.4155959801551965, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.306694030761719, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8592737913131714, "num_tokens": 424688735.0, "step": 11128 }, { "epoch": 1.415723190433787, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.300691604614258, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8709145784378052, "num_tokens": 424732859.0, "step": 11129 }, { "epoch": 1.4158504007123776, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.35267448425293, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8615903854370117, "num_tokens": 424769381.0, "step": 11130 }, { "epoch": 1.415977610990968, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.323963165283203, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8697915077209473, "num_tokens": 424802362.0, "step": 11131 }, { "epoch": 1.4161048212695586, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.294061660766602, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.86651611328125, "num_tokens": 424837897.0, "step": 11132 }, { "epoch": 1.4162320315481491, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.310871124267578, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8807485699653625, "num_tokens": 424878154.0, "step": 11133 }, { "epoch": 1.4163592418267397, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.357800483703613, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8752150535583496, "num_tokens": 424915063.0, "step": 11134 }, { "epoch": 1.41648645210533, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.335330963134766, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8688458204269409, "num_tokens": 424955159.0, "step": 11135 }, { "epoch": 1.4166136623839205, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.29863452911377, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8605927228927612, "num_tokens": 424992573.0, "step": 11136 }, { "epoch": 1.416740872662511, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.354373931884766, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8739597797393799, "num_tokens": 425030452.0, "step": 11137 }, { "epoch": 1.4168680829411016, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.276531219482422, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8828436732292175, "num_tokens": 425067118.0, "step": 11138 }, { "epoch": 1.416995293219692, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.457688331604004, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8601396083831787, "num_tokens": 425101684.0, "step": 11139 }, { "epoch": 1.4171225034982826, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.340351104736328, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8778531551361084, "num_tokens": 425143127.0, "step": 11140 }, { "epoch": 1.4172497137768731, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.372638702392578, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8696649074554443, "num_tokens": 425173912.0, "step": 11141 }, { "epoch": 1.4173769240554637, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.41104507446289, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8739889860153198, "num_tokens": 425210146.0, "step": 11142 }, { "epoch": 1.4175041343340542, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.351070404052734, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8818221688270569, "num_tokens": 425246082.0, "step": 11143 }, { "epoch": 1.4176313446126447, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.348997116088867, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8662127256393433, "num_tokens": 425287470.0, "step": 11144 }, { "epoch": 1.4177585548912353, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.392801284790039, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8659204244613647, "num_tokens": 425326553.0, "step": 11145 }, { "epoch": 1.4178857651698258, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.311541557312012, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8621587157249451, "num_tokens": 425363105.0, "step": 11146 }, { "epoch": 1.4180129754484163, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.344497680664062, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8626202940940857, "num_tokens": 425402513.0, "step": 11147 }, { "epoch": 1.4181401857270068, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.356685638427734, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8694818019866943, "num_tokens": 425440132.0, "step": 11148 }, { "epoch": 1.4182673960055974, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.31722640991211, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8680405020713806, "num_tokens": 425481738.0, "step": 11149 }, { "epoch": 1.418394606284188, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.366801261901855, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8676407933235168, "num_tokens": 425516916.0, "step": 11150 }, { "epoch": 1.4185218165627782, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.323867797851562, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8683958053588867, "num_tokens": 425555951.0, "step": 11151 }, { "epoch": 1.4186490268413687, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.292292594909668, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.86939936876297, "num_tokens": 425595302.0, "step": 11152 }, { "epoch": 1.4187762371199593, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.364471435546875, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8716692924499512, "num_tokens": 425623281.0, "step": 11153 }, { "epoch": 1.4189034473985498, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.322064399719238, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8775341510772705, "num_tokens": 425659124.0, "step": 11154 }, { "epoch": 1.4190306576771403, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.300806999206543, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8581627607345581, "num_tokens": 425695809.0, "step": 11155 }, { "epoch": 1.4191578679557308, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.298258781433105, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8602789640426636, "num_tokens": 425731890.0, "step": 11156 }, { "epoch": 1.4192850782343214, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.373740196228027, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8773826360702515, "num_tokens": 425770899.0, "step": 11157 }, { "epoch": 1.419412288512912, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.307668685913086, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.879640519618988, "num_tokens": 425807812.0, "step": 11158 }, { "epoch": 1.4195394987915024, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.365883827209473, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8649812340736389, "num_tokens": 425848997.0, "step": 11159 }, { "epoch": 1.4196667090700927, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.298775672912598, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8789913058280945, "num_tokens": 425887816.0, "step": 11160 }, { "epoch": 1.4197939193486833, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.344548225402832, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8607784509658813, "num_tokens": 425925488.0, "step": 11161 }, { "epoch": 1.4199211296272738, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.309353828430176, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8753063678741455, "num_tokens": 425965616.0, "step": 11162 }, { "epoch": 1.4200483399058643, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.301901817321777, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8626211881637573, "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": 12.350584983825684, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8655853867530823, "num_tokens": 426042339.0, "step": 11164 }, { "epoch": 1.4203027604630454, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.319855690002441, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8683809041976929, "num_tokens": 426081925.0, "step": 11165 }, { "epoch": 1.420429970741636, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.327783584594727, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8637151122093201, "num_tokens": 426120918.0, "step": 11166 }, { "epoch": 1.4205571810202264, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.344300270080566, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.857901394367218, "num_tokens": 426158711.0, "step": 11167 }, { "epoch": 1.420684391298817, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.274216651916504, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8651472330093384, "num_tokens": 426197034.0, "step": 11168 }, { "epoch": 1.4208116015774075, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.355181694030762, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8621658086776733, "num_tokens": 426239390.0, "step": 11169 }, { "epoch": 1.420938811855998, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.295991897583008, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8586128950119019, "num_tokens": 426285488.0, "step": 11170 }, { "epoch": 1.4210660221345885, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.35375690460205, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8555389642715454, "num_tokens": 426323955.0, "step": 11171 }, { "epoch": 1.421193232413179, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.32222843170166, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8869403004646301, "num_tokens": 426359460.0, "step": 11172 }, { "epoch": 1.4213204426917696, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.379788398742676, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8608508110046387, "num_tokens": 426397104.0, "step": 11173 }, { "epoch": 1.4214476529703601, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.36555290222168, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8774241209030151, "num_tokens": 426441349.0, "step": 11174 }, { "epoch": 1.4215748632489504, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.283026695251465, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8769397735595703, "num_tokens": 426478410.0, "step": 11175 }, { "epoch": 1.421702073527541, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.30377197265625, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8666687607765198, "num_tokens": 426519874.0, "step": 11176 }, { "epoch": 1.4218292838061315, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.310297966003418, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8574198484420776, "num_tokens": 426559970.0, "step": 11177 }, { "epoch": 1.421956494084722, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.321410179138184, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8704240322113037, "num_tokens": 426593897.0, "step": 11178 }, { "epoch": 1.4220837043633126, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.279742240905762, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8686448931694031, "num_tokens": 426633712.0, "step": 11179 }, { "epoch": 1.422210914641903, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.334283828735352, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8836690187454224, "num_tokens": 426669085.0, "step": 11180 }, { "epoch": 1.4223381249204936, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.262487411499023, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8813289403915405, "num_tokens": 426709260.0, "step": 11181 }, { "epoch": 1.4224653351990841, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.347189903259277, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8582737445831299, "num_tokens": 426747747.0, "step": 11182 }, { "epoch": 1.4225925454776747, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.354242324829102, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8560269474983215, "num_tokens": 426788185.0, "step": 11183 }, { "epoch": 1.422719755756265, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.334836959838867, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8705094456672668, "num_tokens": 426828663.0, "step": 11184 }, { "epoch": 1.4228469660348555, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.352560997009277, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8597147464752197, "num_tokens": 426869205.0, "step": 11185 }, { "epoch": 1.422974176313446, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.33285140991211, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8838053941726685, "num_tokens": 426900561.0, "step": 11186 }, { "epoch": 1.4231013865920366, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.370741844177246, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8836427927017212, "num_tokens": 426940618.0, "step": 11187 }, { "epoch": 1.423228596870627, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.307650566101074, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.859993577003479, "num_tokens": 426975509.0, "step": 11188 }, { "epoch": 1.4233558071492176, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.36072826385498, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8772361278533936, "num_tokens": 427011647.0, "step": 11189 }, { "epoch": 1.4234830174278081, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.329522132873535, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8663464784622192, "num_tokens": 427046631.0, "step": 11190 }, { "epoch": 1.4236102277063987, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.358552932739258, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8662204146385193, "num_tokens": 427081364.0, "step": 11191 }, { "epoch": 1.4237374379849892, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.26962947845459, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8704200387001038, "num_tokens": 427118667.0, "step": 11192 }, { "epoch": 1.4238646482635797, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.362460136413574, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8745275139808655, "num_tokens": 427155017.0, "step": 11193 }, { "epoch": 1.4239918585421703, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.330259323120117, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.862006425857544, "num_tokens": 427191741.0, "step": 11194 }, { "epoch": 1.4241190688207608, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.355103492736816, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8732651472091675, "num_tokens": 427226210.0, "step": 11195 }, { "epoch": 1.4242462790993513, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.332340240478516, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.865178108215332, "num_tokens": 427261699.0, "step": 11196 }, { "epoch": 1.4243734893779418, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.348699569702148, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8596597909927368, "num_tokens": 427297158.0, "step": 11197 }, { "epoch": 1.4245006996565324, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.318899154663086, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8542929887771606, "num_tokens": 427338234.0, "step": 11198 }, { "epoch": 1.424627909935123, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.332273483276367, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8625426292419434, "num_tokens": 427376727.0, "step": 11199 }, { "epoch": 1.4247551202137132, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.366421699523926, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8809502124786377, "num_tokens": 427417095.0, "step": 11200 }, { "epoch": 1.4248823304923037, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.315489768981934, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8524664640426636, "num_tokens": 427453243.0, "step": 11201 }, { "epoch": 1.4250095407708943, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.287152290344238, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8720879554748535, "num_tokens": 427490961.0, "step": 11202 }, { "epoch": 1.4251367510494848, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.340036392211914, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8780189752578735, "num_tokens": 427521487.0, "step": 11203 }, { "epoch": 1.4252639613280753, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.301337242126465, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8660838603973389, "num_tokens": 427556607.0, "step": 11204 }, { "epoch": 1.4253911716066658, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.283132553100586, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8696728348731995, "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": 12.340170860290527, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8585758209228516, "num_tokens": 427641337.0, "step": 11206 }, { "epoch": 1.425645592163847, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.288712501525879, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8788430690765381, "num_tokens": 427680235.0, "step": 11207 }, { "epoch": 1.4257728024424374, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.320483207702637, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8691142797470093, "num_tokens": 427719918.0, "step": 11208 }, { "epoch": 1.4259000127210277, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.322892189025879, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8711216449737549, "num_tokens": 427758042.0, "step": 11209 }, { "epoch": 1.4260272229996183, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.35334300994873, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8712899088859558, "num_tokens": 427794591.0, "step": 11210 }, { "epoch": 1.4261544332782088, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.311954498291016, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8767423033714294, "num_tokens": 427828190.0, "step": 11211 }, { "epoch": 1.4262816435567993, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.313767433166504, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8617634177207947, "num_tokens": 427870363.0, "step": 11212 }, { "epoch": 1.4264088538353898, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.300054550170898, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.88332599401474, "num_tokens": 427910037.0, "step": 11213 }, { "epoch": 1.4265360641139804, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.315951347351074, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8682792782783508, "num_tokens": 427945782.0, "step": 11214 }, { "epoch": 1.426663274392571, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.298035621643066, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.876312255859375, "num_tokens": 427984635.0, "step": 11215 }, { "epoch": 1.4267904846711614, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.315648078918457, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8844190835952759, "num_tokens": 428024289.0, "step": 11216 }, { "epoch": 1.426917694949752, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.340195655822754, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.860833466053009, "num_tokens": 428067119.0, "step": 11217 }, { "epoch": 1.4270449052283425, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.266082763671875, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8673697710037231, "num_tokens": 428108150.0, "step": 11218 }, { "epoch": 1.427172115506933, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.299332618713379, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8804181814193726, "num_tokens": 428150839.0, "step": 11219 }, { "epoch": 1.4272993257855235, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.251639366149902, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8573826551437378, "num_tokens": 428194325.0, "step": 11220 }, { "epoch": 1.427426536064114, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.304969787597656, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8684614300727844, "num_tokens": 428230879.0, "step": 11221 }, { "epoch": 1.4275537463427046, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.30174446105957, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8798300623893738, "num_tokens": 428265795.0, "step": 11222 }, { "epoch": 1.4276809566212951, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.306357383728027, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8755784034729004, "num_tokens": 428300413.0, "step": 11223 }, { "epoch": 1.4278081668998854, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.300355911254883, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.8899364471435547, "num_tokens": 428339293.0, "step": 11224 }, { "epoch": 1.427935377178476, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.35708999633789, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8814282417297363, "num_tokens": 428374158.0, "step": 11225 }, { "epoch": 1.4280625874570665, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.33223819732666, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8709467649459839, "num_tokens": 428411715.0, "step": 11226 }, { "epoch": 1.428189797735657, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.361791610717773, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8586193323135376, "num_tokens": 428450000.0, "step": 11227 }, { "epoch": 1.4283170080142475, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.34958553314209, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8761149644851685, "num_tokens": 428484546.0, "step": 11228 }, { "epoch": 1.428444218292838, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.329113006591797, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8704410791397095, "num_tokens": 428527519.0, "step": 11229 }, { "epoch": 1.4285714285714286, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.362269401550293, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.865427553653717, "num_tokens": 428568744.0, "step": 11230 }, { "epoch": 1.4286986388500191, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.340018272399902, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8655734062194824, "num_tokens": 428613050.0, "step": 11231 }, { "epoch": 1.4288258491286097, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.355088233947754, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8667575120925903, "num_tokens": 428647334.0, "step": 11232 }, { "epoch": 1.4289530594072, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.380133628845215, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8760843873023987, "num_tokens": 428681015.0, "step": 11233 }, { "epoch": 1.4290802696857905, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.299271583557129, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8628445863723755, "num_tokens": 428718903.0, "step": 11234 }, { "epoch": 1.429207479964381, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.338238716125488, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8787044286727905, "num_tokens": 428756038.0, "step": 11235 }, { "epoch": 1.4293346902429716, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.312539100646973, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8858141899108887, "num_tokens": 428793908.0, "step": 11236 }, { "epoch": 1.429461900521562, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.323626518249512, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8622575402259827, "num_tokens": 428830924.0, "step": 11237 }, { "epoch": 1.4295891108001526, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.358940124511719, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8664802312850952, "num_tokens": 428869083.0, "step": 11238 }, { "epoch": 1.4297163210787431, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.350465774536133, "learning_rate": 1e-06, "loss": 0.4841, "mean_token_accuracy": 0.8448988199234009, "num_tokens": 428903970.0, "step": 11239 }, { "epoch": 1.4298435313573337, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.3214111328125, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8745971322059631, "num_tokens": 428940710.0, "step": 11240 }, { "epoch": 1.4299707416359242, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.334002494812012, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8727825284004211, "num_tokens": 428981160.0, "step": 11241 }, { "epoch": 1.4300979519145147, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.288381576538086, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8709877729415894, "num_tokens": 429022464.0, "step": 11242 }, { "epoch": 1.4302251621931052, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.385612487792969, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8711928129196167, "num_tokens": 429060964.0, "step": 11243 }, { "epoch": 1.4303523724716958, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.284459114074707, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8710438013076782, "num_tokens": 429101389.0, "step": 11244 }, { "epoch": 1.4304795827502863, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.351574897766113, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.866323709487915, "num_tokens": 429135547.0, "step": 11245 }, { "epoch": 1.4306067930288768, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.314120292663574, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8725395202636719, "num_tokens": 429177504.0, "step": 11246 }, { "epoch": 1.4307340033074674, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.313919067382812, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8739467263221741, "num_tokens": 429217923.0, "step": 11247 }, { "epoch": 1.430861213586058, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.390495300292969, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8698128461837769, "num_tokens": 429251514.0, "step": 11248 }, { "epoch": 1.4309884238646482, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.402693748474121, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8544278144836426, "num_tokens": 429291747.0, "step": 11249 }, { "epoch": 1.4311156341432387, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.285423278808594, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8665622472763062, "num_tokens": 429333480.0, "step": 11250 }, { "epoch": 1.4312428444218293, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.345640182495117, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8808424472808838, "num_tokens": 429366482.0, "step": 11251 }, { "epoch": 1.4313700547004198, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.33647346496582, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8620351552963257, "num_tokens": 429400704.0, "step": 11252 }, { "epoch": 1.4314972649790103, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.37756061553955, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8579173684120178, "num_tokens": 429437463.0, "step": 11253 }, { "epoch": 1.4316244752576008, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.326909065246582, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8596476316452026, "num_tokens": 429474496.0, "step": 11254 }, { "epoch": 1.4317516855361914, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.3461275100708, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.856135904788971, "num_tokens": 429514641.0, "step": 11255 }, { "epoch": 1.431878895814782, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.320998191833496, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8694205284118652, "num_tokens": 429554054.0, "step": 11256 }, { "epoch": 1.4320061060933724, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.386863708496094, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8656694293022156, "num_tokens": 429594004.0, "step": 11257 }, { "epoch": 1.4321333163719627, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.250279426574707, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8757914900779724, "num_tokens": 429633019.0, "step": 11258 }, { "epoch": 1.4322605266505533, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.391425132751465, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8701169490814209, "num_tokens": 429666869.0, "step": 11259 }, { "epoch": 1.4323877369291438, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.293742179870605, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8786699771881104, "num_tokens": 429700049.0, "step": 11260 }, { "epoch": 1.4325149472077343, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.386091232299805, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8669044971466064, "num_tokens": 429737517.0, "step": 11261 }, { "epoch": 1.4326421574863248, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.360304832458496, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.881070077419281, "num_tokens": 429778697.0, "step": 11262 }, { "epoch": 1.4327693677649154, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.365191459655762, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8694415092468262, "num_tokens": 429811883.0, "step": 11263 }, { "epoch": 1.432896578043506, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.309948921203613, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8747557401657104, "num_tokens": 429851748.0, "step": 11264 }, { "epoch": 1.4330237883220964, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.369956016540527, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8593654632568359, "num_tokens": 429891033.0, "step": 11265 }, { "epoch": 1.433150998600687, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.334550857543945, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8710901737213135, "num_tokens": 429935980.0, "step": 11266 }, { "epoch": 1.4332782088792775, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.371435165405273, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8573364019393921, "num_tokens": 429970874.0, "step": 11267 }, { "epoch": 1.433405419157868, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.341634750366211, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8540258407592773, "num_tokens": 430017909.0, "step": 11268 }, { "epoch": 1.4335326294364585, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.378252983093262, "learning_rate": 1e-06, "loss": 0.5166, "mean_token_accuracy": 0.8441678881645203, "num_tokens": 430050641.0, "step": 11269 }, { "epoch": 1.433659839715049, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.288013458251953, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8735946416854858, "num_tokens": 430082985.0, "step": 11270 }, { "epoch": 1.4337870499936396, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.355521202087402, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.869460940361023, "num_tokens": 430122340.0, "step": 11271 }, { "epoch": 1.4339142602722301, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.316228866577148, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8561148643493652, "num_tokens": 430162140.0, "step": 11272 }, { "epoch": 1.4340414705508204, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.303141593933105, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8793423175811768, "num_tokens": 430201672.0, "step": 11273 }, { "epoch": 1.434168680829411, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.34599781036377, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8592699766159058, "num_tokens": 430239760.0, "step": 11274 }, { "epoch": 1.4342958911080015, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.350919723510742, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8521542549133301, "num_tokens": 430280527.0, "step": 11275 }, { "epoch": 1.434423101386592, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.345284461975098, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8646506071090698, "num_tokens": 430312788.0, "step": 11276 }, { "epoch": 1.4345503116651825, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.296934127807617, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8758431673049927, "num_tokens": 430348044.0, "step": 11277 }, { "epoch": 1.434677521943773, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.371527671813965, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8836541771888733, "num_tokens": 430385676.0, "step": 11278 }, { "epoch": 1.4348047322223636, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.35091781616211, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8660286068916321, "num_tokens": 430416976.0, "step": 11279 }, { "epoch": 1.4349319425009541, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.316637992858887, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8799500465393066, "num_tokens": 430460002.0, "step": 11280 }, { "epoch": 1.4350591527795447, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.31586742401123, "learning_rate": 1e-06, "loss": 0.4719, "mean_token_accuracy": 0.8476046323776245, "num_tokens": 430503172.0, "step": 11281 }, { "epoch": 1.435186363058135, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.385010719299316, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8564842939376831, "num_tokens": 430541221.0, "step": 11282 }, { "epoch": 1.4353135733367255, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.313830375671387, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8617798089981079, "num_tokens": 430579527.0, "step": 11283 }, { "epoch": 1.435440783615316, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.320123672485352, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8714389204978943, "num_tokens": 430617263.0, "step": 11284 }, { "epoch": 1.4355679938939065, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.371757507324219, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8821920156478882, "num_tokens": 430649961.0, "step": 11285 }, { "epoch": 1.435695204172497, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.32261848449707, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8632493019104004, "num_tokens": 430682599.0, "step": 11286 }, { "epoch": 1.4358224144510876, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.358509063720703, "learning_rate": 1e-06, "loss": 0.4966, "mean_token_accuracy": 0.8412948846817017, "num_tokens": 430721014.0, "step": 11287 }, { "epoch": 1.4359496247296781, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.32175064086914, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8665766716003418, "num_tokens": 430757014.0, "step": 11288 }, { "epoch": 1.4360768350082687, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.354796409606934, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8696334362030029, "num_tokens": 430796169.0, "step": 11289 }, { "epoch": 1.4362040452868592, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.349936485290527, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8670445084571838, "num_tokens": 430832726.0, "step": 11290 }, { "epoch": 1.4363312555654497, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.31911563873291, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8571152091026306, "num_tokens": 430866343.0, "step": 11291 }, { "epoch": 1.4364584658440402, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.34023380279541, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8740471601486206, "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": 12.29559326171875, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8819873332977295, "num_tokens": 430937499.0, "step": 11293 }, { "epoch": 1.4367128864012213, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.393585205078125, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8566148281097412, "num_tokens": 430974048.0, "step": 11294 }, { "epoch": 1.4368400966798118, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.370176315307617, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8773238062858582, "num_tokens": 431010927.0, "step": 11295 }, { "epoch": 1.4369673069584024, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.376026153564453, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.853599488735199, "num_tokens": 431051731.0, "step": 11296 }, { "epoch": 1.4370945172369929, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.358135223388672, "learning_rate": 1e-06, "loss": 0.3536, "mean_token_accuracy": 0.8831056356430054, "num_tokens": 431087876.0, "step": 11297 }, { "epoch": 1.4372217275155832, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.353260040283203, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8620606064796448, "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": 12.312492370605469, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8689670562744141, "num_tokens": 431168049.0, "step": 11299 }, { "epoch": 1.4374761480727642, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.36929702758789, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8663098812103271, "num_tokens": 431211143.0, "step": 11300 }, { "epoch": 1.4376033583513548, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.324780464172363, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8758980631828308, "num_tokens": 431255747.0, "step": 11301 }, { "epoch": 1.4377305686299453, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.31513786315918, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8717265725135803, "num_tokens": 431296618.0, "step": 11302 }, { "epoch": 1.4378577789085358, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.3493013381958, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8692212104797363, "num_tokens": 431336493.0, "step": 11303 }, { "epoch": 1.4379849891871264, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.374892234802246, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8660742044448853, "num_tokens": 431376200.0, "step": 11304 }, { "epoch": 1.438112199465717, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.37699031829834, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8489400148391724, "num_tokens": 431414015.0, "step": 11305 }, { "epoch": 1.4382394097443074, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.383041381835938, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8642660975456238, "num_tokens": 431453498.0, "step": 11306 }, { "epoch": 1.4383666200228977, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.334378242492676, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8672057390213013, "num_tokens": 431491126.0, "step": 11307 }, { "epoch": 1.4384938303014883, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.298905372619629, "learning_rate": 1e-06, "loss": 0.338, "mean_token_accuracy": 0.8908736705780029, "num_tokens": 431528549.0, "step": 11308 }, { "epoch": 1.4386210405800788, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.395978927612305, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8736270070075989, "num_tokens": 431569435.0, "step": 11309 }, { "epoch": 1.4387482508586693, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.359027862548828, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8734840154647827, "num_tokens": 431606125.0, "step": 11310 }, { "epoch": 1.4388754611372598, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.3336181640625, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.863123893737793, "num_tokens": 431643517.0, "step": 11311 }, { "epoch": 1.4390026714158504, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.329483032226562, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8782260417938232, "num_tokens": 431682058.0, "step": 11312 }, { "epoch": 1.439129881694441, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.319512367248535, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8803110718727112, "num_tokens": 431721241.0, "step": 11313 }, { "epoch": 1.4392570919730314, "ewc_loss": 0.0283203125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 12.397719383239746, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8813738226890564, "num_tokens": 431757809.0, "step": 11314 }, { "epoch": 1.439384302251622, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.286661148071289, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8574718832969666, "num_tokens": 431795931.0, "step": 11315 }, { "epoch": 1.4395115125302125, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.36661434173584, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8773127794265747, "num_tokens": 431830517.0, "step": 11316 }, { "epoch": 1.439638722808803, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.385458946228027, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8708884716033936, "num_tokens": 431866047.0, "step": 11317 }, { "epoch": 1.4397659330873935, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.36420726776123, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.856972336769104, "num_tokens": 431901602.0, "step": 11318 }, { "epoch": 1.439893143365984, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.406143188476562, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8649225234985352, "num_tokens": 431934989.0, "step": 11319 }, { "epoch": 1.4400203536445746, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.368011474609375, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8723167181015015, "num_tokens": 431968664.0, "step": 11320 }, { "epoch": 1.4401475639231651, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.336213111877441, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8797341585159302, "num_tokens": 432008025.0, "step": 11321 }, { "epoch": 1.4402747742017554, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.372064590454102, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.873348593711853, "num_tokens": 432043526.0, "step": 11322 }, { "epoch": 1.440401984480346, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.302506446838379, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8697267174720764, "num_tokens": 432085175.0, "step": 11323 }, { "epoch": 1.4405291947589365, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.355356216430664, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8601241707801819, "num_tokens": 432122379.0, "step": 11324 }, { "epoch": 1.440656405037527, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.36564826965332, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8713427186012268, "num_tokens": 432157690.0, "step": 11325 }, { "epoch": 1.4407836153161175, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.332160949707031, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8766202926635742, "num_tokens": 432188885.0, "step": 11326 }, { "epoch": 1.440910825594708, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.378310203552246, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8686774969100952, "num_tokens": 432228804.0, "step": 11327 }, { "epoch": 1.4410380358732986, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.323713302612305, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8509085774421692, "num_tokens": 432264150.0, "step": 11328 }, { "epoch": 1.4411652461518891, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.379318237304688, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8623283505439758, "num_tokens": 432306759.0, "step": 11329 }, { "epoch": 1.4412924564304797, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.319268226623535, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8788586854934692, "num_tokens": 432346928.0, "step": 11330 }, { "epoch": 1.44141966670907, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.263429641723633, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.87486732006073, "num_tokens": 432386630.0, "step": 11331 }, { "epoch": 1.4415468769876605, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.393667221069336, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8564894199371338, "num_tokens": 432427331.0, "step": 11332 }, { "epoch": 1.441674087266251, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.379179000854492, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8614215850830078, "num_tokens": 432462901.0, "step": 11333 }, { "epoch": 1.4418012975448415, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.343269348144531, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8662612438201904, "num_tokens": 432494232.0, "step": 11334 }, { "epoch": 1.441928507823432, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.324939727783203, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8808218836784363, "num_tokens": 432534276.0, "step": 11335 }, { "epoch": 1.4420557181020226, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.327187538146973, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8603307604789734, "num_tokens": 432577337.0, "step": 11336 }, { "epoch": 1.4421829283806131, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.390759468078613, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8606273531913757, "num_tokens": 432614290.0, "step": 11337 }, { "epoch": 1.4423101386592037, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.388659477233887, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8698558807373047, "num_tokens": 432648558.0, "step": 11338 }, { "epoch": 1.4424373489377942, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.369475364685059, "learning_rate": 1e-06, "loss": 0.4958, "mean_token_accuracy": 0.8418742418289185, "num_tokens": 432684243.0, "step": 11339 }, { "epoch": 1.4425645592163847, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.37298583984375, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8676807880401611, "num_tokens": 432719355.0, "step": 11340 }, { "epoch": 1.4426917694949752, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.417943954467773, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8682419061660767, "num_tokens": 432758675.0, "step": 11341 }, { "epoch": 1.4428189797735658, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.30386734008789, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8655742406845093, "num_tokens": 432801677.0, "step": 11342 }, { "epoch": 1.4429461900521563, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.421003341674805, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8660258054733276, "num_tokens": 432836299.0, "step": 11343 }, { "epoch": 1.4430734003307468, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.40791130065918, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.866680383682251, "num_tokens": 432868838.0, "step": 11344 }, { "epoch": 1.4432006106093374, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.33419132232666, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8695162534713745, "num_tokens": 432904895.0, "step": 11345 }, { "epoch": 1.4433278208879279, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.315032958984375, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8627846240997314, "num_tokens": 432944785.0, "step": 11346 }, { "epoch": 1.4434550311665182, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.367631912231445, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8586710691452026, "num_tokens": 432975645.0, "step": 11347 }, { "epoch": 1.4435822414451087, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.33161735534668, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8631446361541748, "num_tokens": 433016887.0, "step": 11348 }, { "epoch": 1.4437094517236992, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.298029899597168, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8693456053733826, "num_tokens": 433053666.0, "step": 11349 }, { "epoch": 1.4438366620022898, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.321979522705078, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8636452555656433, "num_tokens": 433101131.0, "step": 11350 }, { "epoch": 1.4439638722808803, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.351005554199219, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8751450181007385, "num_tokens": 433132535.0, "step": 11351 }, { "epoch": 1.4440910825594708, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.374817848205566, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8539924621582031, "num_tokens": 433168229.0, "step": 11352 }, { "epoch": 1.4442182928380614, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.359310150146484, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8733389973640442, "num_tokens": 433206094.0, "step": 11353 }, { "epoch": 1.4443455031166519, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.359106063842773, "learning_rate": 1e-06, "loss": 0.4574, "mean_token_accuracy": 0.855441689491272, "num_tokens": 433241038.0, "step": 11354 }, { "epoch": 1.4444727133952424, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.365923881530762, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.860837459564209, "num_tokens": 433274674.0, "step": 11355 }, { "epoch": 1.4445999236738327, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.37318229675293, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8716821074485779, "num_tokens": 433314511.0, "step": 11356 }, { "epoch": 1.4447271339524232, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.338339805603027, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8697401881217957, "num_tokens": 433349338.0, "step": 11357 }, { "epoch": 1.4448543442310138, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.344433784484863, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8784334659576416, "num_tokens": 433382609.0, "step": 11358 }, { "epoch": 1.4449815545096043, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.408462524414062, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8822924494743347, "num_tokens": 433411553.0, "step": 11359 }, { "epoch": 1.4451087647881948, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.351446151733398, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8661425709724426, "num_tokens": 433450571.0, "step": 11360 }, { "epoch": 1.4452359750667854, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.409006118774414, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8630151748657227, "num_tokens": 433486994.0, "step": 11361 }, { "epoch": 1.445363185345376, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.355513572692871, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8804450631141663, "num_tokens": 433521407.0, "step": 11362 }, { "epoch": 1.4454903956239664, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.398344993591309, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8793251514434814, "num_tokens": 433561338.0, "step": 11363 }, { "epoch": 1.445617605902557, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.384346961975098, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8751726150512695, "num_tokens": 433598892.0, "step": 11364 }, { "epoch": 1.4457448161811475, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.397554397583008, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8575404286384583, "num_tokens": 433641423.0, "step": 11365 }, { "epoch": 1.445872026459738, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.372854232788086, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8709849715232849, "num_tokens": 433677385.0, "step": 11366 }, { "epoch": 1.4459992367383285, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.307531356811523, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8660703301429749, "num_tokens": 433716901.0, "step": 11367 }, { "epoch": 1.446126447016919, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.370260238647461, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8639605045318604, "num_tokens": 433755099.0, "step": 11368 }, { "epoch": 1.4462536572955096, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.362268447875977, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8648920059204102, "num_tokens": 433794012.0, "step": 11369 }, { "epoch": 1.4463808675741001, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.330809593200684, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8714215755462646, "num_tokens": 433831965.0, "step": 11370 }, { "epoch": 1.4465080778526904, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.371525764465332, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8688453435897827, "num_tokens": 433870819.0, "step": 11371 }, { "epoch": 1.446635288131281, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.344057083129883, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.862121045589447, "num_tokens": 433909197.0, "step": 11372 }, { "epoch": 1.4467624984098715, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.367688179016113, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8749780058860779, "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": 12.350055694580078, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8900238275527954, "num_tokens": 433986239.0, "step": 11374 }, { "epoch": 1.4470169189670525, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.370455741882324, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8732684850692749, "num_tokens": 434022557.0, "step": 11375 }, { "epoch": 1.447144129245643, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.35557746887207, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8724441528320312, "num_tokens": 434059258.0, "step": 11376 }, { "epoch": 1.4472713395242336, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.339705467224121, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8548638820648193, "num_tokens": 434094872.0, "step": 11377 }, { "epoch": 1.4473985498028241, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.362902641296387, "learning_rate": 1e-06, "loss": 0.4849, "mean_token_accuracy": 0.8458596467971802, "num_tokens": 434134842.0, "step": 11378 }, { "epoch": 1.4475257600814146, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.335067749023438, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8802711963653564, "num_tokens": 434173152.0, "step": 11379 }, { "epoch": 1.447652970360005, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.381516456604004, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8699744939804077, "num_tokens": 434208318.0, "step": 11380 }, { "epoch": 1.4477801806385955, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.420517921447754, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8663531541824341, "num_tokens": 434243693.0, "step": 11381 }, { "epoch": 1.447907390917186, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.331243515014648, "learning_rate": 1e-06, "loss": 0.462, "mean_token_accuracy": 0.8562005758285522, "num_tokens": 434283172.0, "step": 11382 }, { "epoch": 1.4480346011957765, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.4222412109375, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8698124885559082, "num_tokens": 434318789.0, "step": 11383 }, { "epoch": 1.448161811474367, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.347845077514648, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8600411415100098, "num_tokens": 434358943.0, "step": 11384 }, { "epoch": 1.4482890217529576, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.376754760742188, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8697118759155273, "num_tokens": 434398264.0, "step": 11385 }, { "epoch": 1.4484162320315481, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.330850601196289, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8775939345359802, "num_tokens": 434433602.0, "step": 11386 }, { "epoch": 1.4485434423101387, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.348220825195312, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8615226745605469, "num_tokens": 434480001.0, "step": 11387 }, { "epoch": 1.4486706525887292, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.334796905517578, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8839002847671509, "num_tokens": 434519285.0, "step": 11388 }, { "epoch": 1.4487978628673197, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.420228004455566, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8685773015022278, "num_tokens": 434556118.0, "step": 11389 }, { "epoch": 1.4489250731459102, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.415986061096191, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8530014753341675, "num_tokens": 434595350.0, "step": 11390 }, { "epoch": 1.4490522834245008, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.318073272705078, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8657410740852356, "num_tokens": 434632988.0, "step": 11391 }, { "epoch": 1.4491794937030913, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.384077072143555, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8752221465110779, "num_tokens": 434673040.0, "step": 11392 }, { "epoch": 1.4493067039816818, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.417845726013184, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8832892179489136, "num_tokens": 434712201.0, "step": 11393 }, { "epoch": 1.4494339142602723, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.330135345458984, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8824630975723267, "num_tokens": 434756676.0, "step": 11394 }, { "epoch": 1.4495611245388629, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.346627235412598, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8726409673690796, "num_tokens": 434797890.0, "step": 11395 }, { "epoch": 1.4496883348174532, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.372353553771973, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8822386264801025, "num_tokens": 434836933.0, "step": 11396 }, { "epoch": 1.4498155450960437, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.43215560913086, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8818813562393188, "num_tokens": 434868098.0, "step": 11397 }, { "epoch": 1.4499427553746342, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.378292083740234, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8647708892822266, "num_tokens": 434907239.0, "step": 11398 }, { "epoch": 1.4500699656532248, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.392168045043945, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8748799562454224, "num_tokens": 434945672.0, "step": 11399 }, { "epoch": 1.4501971759318153, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.361517906188965, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.858539342880249, "num_tokens": 434981693.0, "step": 11400 }, { "epoch": 1.4503243862104058, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.334229469299316, "learning_rate": 1e-06, "loss": 0.3339, "mean_token_accuracy": 0.8907763957977295, "num_tokens": 435012452.0, "step": 11401 }, { "epoch": 1.4504515964889964, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.368022918701172, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8626961708068848, "num_tokens": 435052711.0, "step": 11402 }, { "epoch": 1.4505788067675869, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.388906478881836, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8666254878044128, "num_tokens": 435092766.0, "step": 11403 }, { "epoch": 1.4507060170461774, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.360265731811523, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8543802499771118, "num_tokens": 435132093.0, "step": 11404 }, { "epoch": 1.4508332273247677, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.399560928344727, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8664301633834839, "num_tokens": 435169071.0, "step": 11405 }, { "epoch": 1.4509604376033582, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.432534217834473, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8794969320297241, "num_tokens": 435202665.0, "step": 11406 }, { "epoch": 1.4510876478819488, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.353907585144043, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8831324577331543, "num_tokens": 435234719.0, "step": 11407 }, { "epoch": 1.4512148581605393, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.481929779052734, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8830649256706238, "num_tokens": 435267090.0, "step": 11408 }, { "epoch": 1.4513420684391298, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.400973320007324, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8764864206314087, "num_tokens": 435307243.0, "step": 11409 }, { "epoch": 1.4514692787177204, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.373230934143066, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.876151442527771, "num_tokens": 435347507.0, "step": 11410 }, { "epoch": 1.4515964889963109, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.393513679504395, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8603289127349854, "num_tokens": 435383499.0, "step": 11411 }, { "epoch": 1.4517236992749014, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.405101776123047, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8690487146377563, "num_tokens": 435423500.0, "step": 11412 }, { "epoch": 1.451850909553492, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.44692325592041, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8735374212265015, "num_tokens": 435457075.0, "step": 11413 }, { "epoch": 1.4519781198320825, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.390741348266602, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8676536083221436, "num_tokens": 435498933.0, "step": 11414 }, { "epoch": 1.452105330110673, "ewc_loss": 0.028564453125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 12.424177169799805, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8606897592544556, "num_tokens": 435535949.0, "step": 11415 }, { "epoch": 1.4522325403892635, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.4324951171875, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8632555603981018, "num_tokens": 435565252.0, "step": 11416 }, { "epoch": 1.452359750667854, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.374369621276855, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.849943220615387, "num_tokens": 435604663.0, "step": 11417 }, { "epoch": 1.4524869609464446, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.379716873168945, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8636627197265625, "num_tokens": 435640883.0, "step": 11418 }, { "epoch": 1.4526141712250351, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.383075714111328, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8718880414962769, "num_tokens": 435682787.0, "step": 11419 }, { "epoch": 1.4527413815036254, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.459656715393066, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8666306138038635, "num_tokens": 435719982.0, "step": 11420 }, { "epoch": 1.452868591782216, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.360721588134766, "learning_rate": 1e-06, "loss": 0.4776, "mean_token_accuracy": 0.8520521521568298, "num_tokens": 435755153.0, "step": 11421 }, { "epoch": 1.4529958020608065, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.421791076660156, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8780450820922852, "num_tokens": 435797513.0, "step": 11422 }, { "epoch": 1.453123012339397, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.389863967895508, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8687959909439087, "num_tokens": 435836835.0, "step": 11423 }, { "epoch": 1.4532502226179875, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.395905494689941, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8600026369094849, "num_tokens": 435875124.0, "step": 11424 }, { "epoch": 1.453377432896578, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.422895431518555, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.881014347076416, "num_tokens": 435914993.0, "step": 11425 }, { "epoch": 1.4535046431751686, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.405421257019043, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8767955899238586, "num_tokens": 435950916.0, "step": 11426 }, { "epoch": 1.4536318534537591, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.374439239501953, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8637529611587524, "num_tokens": 435991986.0, "step": 11427 }, { "epoch": 1.4537590637323496, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.472665786743164, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.877282977104187, "num_tokens": 436024603.0, "step": 11428 }, { "epoch": 1.45388627401094, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.345602989196777, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8649831414222717, "num_tokens": 436065540.0, "step": 11429 }, { "epoch": 1.4540134842895305, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.388206481933594, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8552191257476807, "num_tokens": 436104394.0, "step": 11430 }, { "epoch": 1.454140694568121, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.356287956237793, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8552771210670471, "num_tokens": 436139977.0, "step": 11431 }, { "epoch": 1.4542679048467115, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.366997718811035, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8839722275733948, "num_tokens": 436178047.0, "step": 11432 }, { "epoch": 1.454395115125302, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.346348762512207, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8677977323532104, "num_tokens": 436218331.0, "step": 11433 }, { "epoch": 1.4545223254038926, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.342622756958008, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8660922050476074, "num_tokens": 436255663.0, "step": 11434 }, { "epoch": 1.4546495356824831, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.341630935668945, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8684656620025635, "num_tokens": 436301328.0, "step": 11435 }, { "epoch": 1.4547767459610736, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.40736198425293, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8768875598907471, "num_tokens": 436336487.0, "step": 11436 }, { "epoch": 1.4549039562396642, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.376935958862305, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8714140057563782, "num_tokens": 436377629.0, "step": 11437 }, { "epoch": 1.4550311665182547, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.338671684265137, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8823446035385132, "num_tokens": 436421590.0, "step": 11438 }, { "epoch": 1.4551583767968452, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.390093803405762, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8618866205215454, "num_tokens": 436461172.0, "step": 11439 }, { "epoch": 1.4552855870754358, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.347938537597656, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8775894045829773, "num_tokens": 436497536.0, "step": 11440 }, { "epoch": 1.4554127973540263, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.30871295928955, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8651856184005737, "num_tokens": 436543906.0, "step": 11441 }, { "epoch": 1.4555400076326168, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.369282722473145, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8703449964523315, "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": 12.349569320678711, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8610965013504028, "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": 12.349980354309082, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8571848273277283, "num_tokens": 436654176.0, "step": 11444 }, { "epoch": 1.4559216384683882, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.320843696594238, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8805059194564819, "num_tokens": 436685507.0, "step": 11445 }, { "epoch": 1.4560488487469787, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.394587516784668, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8771698474884033, "num_tokens": 436729881.0, "step": 11446 }, { "epoch": 1.4561760590255692, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.345312118530273, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8676546812057495, "num_tokens": 436772355.0, "step": 11447 }, { "epoch": 1.4563032693041598, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.401069641113281, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8624938130378723, "num_tokens": 436808701.0, "step": 11448 }, { "epoch": 1.4564304795827503, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.37031078338623, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8784358501434326, "num_tokens": 436845441.0, "step": 11449 }, { "epoch": 1.4565576898613408, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.442530632019043, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8709637522697449, "num_tokens": 436885316.0, "step": 11450 }, { "epoch": 1.4566849001399313, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.359465599060059, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8634979724884033, "num_tokens": 436918315.0, "step": 11451 }, { "epoch": 1.4568121104185219, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.437997817993164, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8608973026275635, "num_tokens": 436956039.0, "step": 11452 }, { "epoch": 1.4569393206971124, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.356376647949219, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8544725775718689, "num_tokens": 436991274.0, "step": 11453 }, { "epoch": 1.4570665309757027, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.393486976623535, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8799200057983398, "num_tokens": 437029963.0, "step": 11454 }, { "epoch": 1.4571937412542932, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.359240531921387, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8620595335960388, "num_tokens": 437067030.0, "step": 11455 }, { "epoch": 1.4573209515328838, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.377629280090332, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8758097290992737, "num_tokens": 437103363.0, "step": 11456 }, { "epoch": 1.4574481618114743, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.413772583007812, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8752104640007019, "num_tokens": 437131574.0, "step": 11457 }, { "epoch": 1.4575753720900648, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.37454605102539, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8614485263824463, "num_tokens": 437169537.0, "step": 11458 }, { "epoch": 1.4577025823686554, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.37116813659668, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8741937875747681, "num_tokens": 437208025.0, "step": 11459 }, { "epoch": 1.4578297926472459, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.38037395477295, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.879468560218811, "num_tokens": 437246642.0, "step": 11460 }, { "epoch": 1.4579570029258364, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.36031436920166, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8663932085037231, "num_tokens": 437285274.0, "step": 11461 }, { "epoch": 1.458084213204427, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.360282897949219, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8859145641326904, "num_tokens": 437325865.0, "step": 11462 }, { "epoch": 1.4582114234830175, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.381930351257324, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8748828172683716, "num_tokens": 437364093.0, "step": 11463 }, { "epoch": 1.458338633761608, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.327031135559082, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8559376001358032, "num_tokens": 437402413.0, "step": 11464 }, { "epoch": 1.4584658440401985, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.3799409866333, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8551459312438965, "num_tokens": 437442583.0, "step": 11465 }, { "epoch": 1.458593054318789, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.34731674194336, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.86858069896698, "num_tokens": 437485682.0, "step": 11466 }, { "epoch": 1.4587202645973796, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.372233390808105, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.858810305595398, "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": 12.389799118041992, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.860315203666687, "num_tokens": 437563590.0, "step": 11468 }, { "epoch": 1.4589746851545604, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.351475715637207, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8851184248924255, "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": 12.370742797851562, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.856535792350769, "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": 12.358328819274902, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8763774633407593, "num_tokens": 437680846.0, "step": 11471 }, { "epoch": 1.459356315990332, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.36773681640625, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8706752061843872, "num_tokens": 437722156.0, "step": 11472 }, { "epoch": 1.4594835262689225, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.380796432495117, "learning_rate": 1e-06, "loss": 0.4898, "mean_token_accuracy": 0.8452991843223572, "num_tokens": 437766905.0, "step": 11473 }, { "epoch": 1.459610736547513, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.318930625915527, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8679406642913818, "num_tokens": 437801425.0, "step": 11474 }, { "epoch": 1.4597379468261036, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.394678115844727, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8585391640663147, "num_tokens": 437838126.0, "step": 11475 }, { "epoch": 1.4598651571046941, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.359525680541992, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8630362749099731, "num_tokens": 437876597.0, "step": 11476 }, { "epoch": 1.4599923673832846, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.372276306152344, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8802036643028259, "num_tokens": 437911852.0, "step": 11477 }, { "epoch": 1.460119577661875, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.319554328918457, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8696962594985962, "num_tokens": 437953418.0, "step": 11478 }, { "epoch": 1.4602467879404655, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.349775314331055, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8745729923248291, "num_tokens": 437989561.0, "step": 11479 }, { "epoch": 1.460373998219056, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.384906768798828, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8773839473724365, "num_tokens": 438033464.0, "step": 11480 }, { "epoch": 1.4605012084976465, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.305865287780762, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8720723390579224, "num_tokens": 438068676.0, "step": 11481 }, { "epoch": 1.460628418776237, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.407461166381836, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.85063236951828, "num_tokens": 438107867.0, "step": 11482 }, { "epoch": 1.4607556290548276, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.338863372802734, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8851057887077332, "num_tokens": 438144256.0, "step": 11483 }, { "epoch": 1.4608828393334181, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.33819580078125, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8789557814598083, "num_tokens": 438183458.0, "step": 11484 }, { "epoch": 1.4610100496120086, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.401049613952637, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8560471534729004, "num_tokens": 438216025.0, "step": 11485 }, { "epoch": 1.4611372598905992, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.374802589416504, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8786513805389404, "num_tokens": 438258185.0, "step": 11486 }, { "epoch": 1.4612644701691897, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.2908353805542, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8643190860748291, "num_tokens": 438298189.0, "step": 11487 }, { "epoch": 1.4613916804477802, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.452526092529297, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8571420311927795, "num_tokens": 438336593.0, "step": 11488 }, { "epoch": 1.4615188907263708, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.367376327514648, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8754161596298218, "num_tokens": 438374110.0, "step": 11489 }, { "epoch": 1.4616461010049613, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.379976272583008, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.8521469831466675, "num_tokens": 438413062.0, "step": 11490 }, { "epoch": 1.4617733112835518, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.452173233032227, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8711166977882385, "num_tokens": 438452055.0, "step": 11491 }, { "epoch": 1.4619005215621423, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.355879783630371, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8708949089050293, "num_tokens": 438494224.0, "step": 11492 }, { "epoch": 1.4620277318407329, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.433305740356445, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8662348985671997, "num_tokens": 438529790.0, "step": 11493 }, { "epoch": 1.4621549421193232, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.336221694946289, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8722609281539917, "num_tokens": 438571371.0, "step": 11494 }, { "epoch": 1.4622821523979137, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.36269474029541, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.854703426361084, "num_tokens": 438612383.0, "step": 11495 }, { "epoch": 1.4624093626765042, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.399503707885742, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8589417338371277, "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": 12.407535552978516, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8572914600372314, "num_tokens": 438689565.0, "step": 11497 }, { "epoch": 1.4626637832336853, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.412336349487305, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8695273399353027, "num_tokens": 438727753.0, "step": 11498 }, { "epoch": 1.4627909935122758, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.372523307800293, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8666380047798157, "num_tokens": 438764667.0, "step": 11499 }, { "epoch": 1.4629182037908663, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.404037475585938, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8743876218795776, "num_tokens": 438801739.0, "step": 11500 }, { "epoch": 1.4630454140694569, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.37275505065918, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8700201511383057, "num_tokens": 438840028.0, "step": 11501 }, { "epoch": 1.4631726243480474, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.420567512512207, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8659083247184753, "num_tokens": 438880632.0, "step": 11502 }, { "epoch": 1.4632998346266377, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.413273811340332, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8660663962364197, "num_tokens": 438923040.0, "step": 11503 }, { "epoch": 1.4634270449052282, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.457974433898926, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8690775632858276, "num_tokens": 438958597.0, "step": 11504 }, { "epoch": 1.4635542551838188, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.393181800842285, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8576620817184448, "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": 12.462380409240723, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.87717604637146, "num_tokens": 439030084.0, "step": 11506 }, { "epoch": 1.4638086757409998, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.485586166381836, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8709207773208618, "num_tokens": 439067770.0, "step": 11507 }, { "epoch": 1.4639358860195903, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.407251358032227, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8715603351593018, "num_tokens": 439102476.0, "step": 11508 }, { "epoch": 1.4640630962981809, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.367119789123535, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8785278797149658, "num_tokens": 439138660.0, "step": 11509 }, { "epoch": 1.4641903065767714, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.408863067626953, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8761901259422302, "num_tokens": 439181690.0, "step": 11510 }, { "epoch": 1.464317516855362, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.443305015563965, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8748478293418884, "num_tokens": 439218825.0, "step": 11511 }, { "epoch": 1.4644447271339525, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.423993110656738, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8735153079032898, "num_tokens": 439253505.0, "step": 11512 }, { "epoch": 1.464571937412543, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.378207206726074, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8632699251174927, "num_tokens": 439292764.0, "step": 11513 }, { "epoch": 1.4646991476911335, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.385178565979004, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8776482343673706, "num_tokens": 439335269.0, "step": 11514 }, { "epoch": 1.464826357969724, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.411539077758789, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8703831434249878, "num_tokens": 439371977.0, "step": 11515 }, { "epoch": 1.4649535682483146, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.394668579101562, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8686448931694031, "num_tokens": 439412710.0, "step": 11516 }, { "epoch": 1.465080778526905, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.433074951171875, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8685077428817749, "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": 12.366145133972168, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8632786273956299, "num_tokens": 439495143.0, "step": 11518 }, { "epoch": 1.465335199084086, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.419057846069336, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8573477268218994, "num_tokens": 439532721.0, "step": 11519 }, { "epoch": 1.4654624093626765, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.413483619689941, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.878078043460846, "num_tokens": 439567938.0, "step": 11520 }, { "epoch": 1.465589619641267, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.478330612182617, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8772273063659668, "num_tokens": 439595720.0, "step": 11521 }, { "epoch": 1.4657168299198575, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.424095153808594, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.869059681892395, "num_tokens": 439634091.0, "step": 11522 }, { "epoch": 1.465844040198448, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.383646965026855, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8576375246047974, "num_tokens": 439671978.0, "step": 11523 }, { "epoch": 1.4659712504770386, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.459002494812012, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8642606139183044, "num_tokens": 439707748.0, "step": 11524 }, { "epoch": 1.466098460755629, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.42342472076416, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8757637739181519, "num_tokens": 439744060.0, "step": 11525 }, { "epoch": 1.4662256710342196, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.344744682312012, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8611873388290405, "num_tokens": 439789277.0, "step": 11526 }, { "epoch": 1.46635288131281, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.410840034484863, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8857482671737671, "num_tokens": 439827524.0, "step": 11527 }, { "epoch": 1.4664800915914005, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.412487030029297, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8658608198165894, "num_tokens": 439865181.0, "step": 11528 }, { "epoch": 1.466607301869991, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.400897026062012, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8612971305847168, "num_tokens": 439900071.0, "step": 11529 }, { "epoch": 1.4667345121485815, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.43200397491455, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8683971762657166, "num_tokens": 439940839.0, "step": 11530 }, { "epoch": 1.466861722427172, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.392760276794434, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8654021620750427, "num_tokens": 439978908.0, "step": 11531 }, { "epoch": 1.4669889327057626, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.376081466674805, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8683224320411682, "num_tokens": 440013448.0, "step": 11532 }, { "epoch": 1.467116142984353, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.453313827514648, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8719559907913208, "num_tokens": 440048377.0, "step": 11533 }, { "epoch": 1.4672433532629436, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.397521018981934, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8631321787834167, "num_tokens": 440089806.0, "step": 11534 }, { "epoch": 1.4673705635415342, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.411540985107422, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.875035285949707, "num_tokens": 440124209.0, "step": 11535 }, { "epoch": 1.4674977738201247, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.394227027893066, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8696292638778687, "num_tokens": 440158349.0, "step": 11536 }, { "epoch": 1.4676249840987152, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.40682601928711, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.871546745300293, "num_tokens": 440195220.0, "step": 11537 }, { "epoch": 1.4677521943773058, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.382002830505371, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.869835615158081, "num_tokens": 440235158.0, "step": 11538 }, { "epoch": 1.4678794046558963, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.457810401916504, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8643747568130493, "num_tokens": 440271879.0, "step": 11539 }, { "epoch": 1.4680066149344868, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.394742012023926, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.86724853515625, "num_tokens": 440306454.0, "step": 11540 }, { "epoch": 1.4681338252130773, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.40722370147705, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8565990924835205, "num_tokens": 440341185.0, "step": 11541 }, { "epoch": 1.4682610354916679, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.397554397583008, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8675169944763184, "num_tokens": 440379168.0, "step": 11542 }, { "epoch": 1.4683882457702582, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.407035827636719, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8709156513214111, "num_tokens": 440417945.0, "step": 11543 }, { "epoch": 1.4685154560488487, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.332069396972656, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8537607192993164, "num_tokens": 440457864.0, "step": 11544 }, { "epoch": 1.4686426663274392, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.417635917663574, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8642797470092773, "num_tokens": 440491709.0, "step": 11545 }, { "epoch": 1.4687698766060298, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.396000862121582, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8746931552886963, "num_tokens": 440533570.0, "step": 11546 }, { "epoch": 1.4688970868846203, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.390643119812012, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8744828104972839, "num_tokens": 440568360.0, "step": 11547 }, { "epoch": 1.4690242971632108, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.374838829040527, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8794687986373901, "num_tokens": 440606848.0, "step": 11548 }, { "epoch": 1.4691515074418013, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.415887832641602, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.881685197353363, "num_tokens": 440638254.0, "step": 11549 }, { "epoch": 1.4692787177203919, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.413457870483398, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8746827244758606, "num_tokens": 440674122.0, "step": 11550 }, { "epoch": 1.4694059279989824, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.418052673339844, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8836218118667603, "num_tokens": 440718362.0, "step": 11551 }, { "epoch": 1.4695331382775727, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.401461601257324, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8759849071502686, "num_tokens": 440755971.0, "step": 11552 }, { "epoch": 1.4696603485561632, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.360405921936035, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8666634559631348, "num_tokens": 440794328.0, "step": 11553 }, { "epoch": 1.4697875588347538, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.388883590698242, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8562366962432861, "num_tokens": 440833383.0, "step": 11554 }, { "epoch": 1.4699147691133443, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.37614631652832, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8754458427429199, "num_tokens": 440875502.0, "step": 11555 }, { "epoch": 1.4700419793919348, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.40606689453125, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8722202777862549, "num_tokens": 440915259.0, "step": 11556 }, { "epoch": 1.4701691896705253, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.335653305053711, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8664309978485107, "num_tokens": 440951660.0, "step": 11557 }, { "epoch": 1.4702963999491159, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.436573028564453, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8682107925415039, "num_tokens": 440992533.0, "step": 11558 }, { "epoch": 1.4704236102277064, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.35616683959961, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8574220538139343, "num_tokens": 441027360.0, "step": 11559 }, { "epoch": 1.470550820506297, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.422396659851074, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8755525350570679, "num_tokens": 441071746.0, "step": 11560 }, { "epoch": 1.4706780307848875, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.452824592590332, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8676564693450928, "num_tokens": 441105378.0, "step": 11561 }, { "epoch": 1.470805241063478, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.447616577148438, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8694822788238525, "num_tokens": 441150000.0, "step": 11562 }, { "epoch": 1.4709324513420685, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.426215171813965, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8609696626663208, "num_tokens": 441188426.0, "step": 11563 }, { "epoch": 1.471059661620659, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.405156135559082, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.87558913230896, "num_tokens": 441226028.0, "step": 11564 }, { "epoch": 1.4711868718992496, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.41563606262207, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8701648712158203, "num_tokens": 441265788.0, "step": 11565 }, { "epoch": 1.47131408217784, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.46364974975586, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8779375553131104, "num_tokens": 441297280.0, "step": 11566 }, { "epoch": 1.4714412924564304, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.418575286865234, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8682353496551514, "num_tokens": 441331229.0, "step": 11567 }, { "epoch": 1.471568502735021, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.420690536499023, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8726469278335571, "num_tokens": 441371064.0, "step": 11568 }, { "epoch": 1.4716957130136115, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.412694931030273, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8504947423934937, "num_tokens": 441405706.0, "step": 11569 }, { "epoch": 1.471822923292202, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.42780590057373, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8604105710983276, "num_tokens": 441445966.0, "step": 11570 }, { "epoch": 1.4719501335707925, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.414046287536621, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8530228734016418, "num_tokens": 441478916.0, "step": 11571 }, { "epoch": 1.472077343849383, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.422595977783203, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8653157949447632, "num_tokens": 441519245.0, "step": 11572 }, { "epoch": 1.4722045541279736, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.436132431030273, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8613181114196777, "num_tokens": 441557485.0, "step": 11573 }, { "epoch": 1.472331764406564, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.37649154663086, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8788790702819824, "num_tokens": 441596042.0, "step": 11574 }, { "epoch": 1.4724589746851546, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.438199043273926, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8784598112106323, "num_tokens": 441634391.0, "step": 11575 }, { "epoch": 1.472586184963745, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.403336524963379, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8612152338027954, "num_tokens": 441672201.0, "step": 11576 }, { "epoch": 1.4727133952423355, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.407346725463867, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8671663403511047, "num_tokens": 441713350.0, "step": 11577 }, { "epoch": 1.472840605520926, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.416459083557129, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8674274682998657, "num_tokens": 441754677.0, "step": 11578 }, { "epoch": 1.4729678157995165, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.363741874694824, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8589649200439453, "num_tokens": 441797037.0, "step": 11579 }, { "epoch": 1.473095026078107, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.440925598144531, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8675095438957214, "num_tokens": 441831661.0, "step": 11580 }, { "epoch": 1.4732222363566976, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.42360782623291, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8722441792488098, "num_tokens": 441876933.0, "step": 11581 }, { "epoch": 1.473349446635288, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.325600624084473, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8902698755264282, "num_tokens": 441918015.0, "step": 11582 }, { "epoch": 1.4734766569138786, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.448802947998047, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8638782501220703, "num_tokens": 441956573.0, "step": 11583 }, { "epoch": 1.4736038671924692, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.390731811523438, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8820079565048218, "num_tokens": 442002064.0, "step": 11584 }, { "epoch": 1.4737310774710597, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.41862678527832, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8764723539352417, "num_tokens": 442040122.0, "step": 11585 }, { "epoch": 1.4738582877496502, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.424301147460938, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8712667226791382, "num_tokens": 442079698.0, "step": 11586 }, { "epoch": 1.4739854980282407, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.423357009887695, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8818421363830566, "num_tokens": 442118252.0, "step": 11587 }, { "epoch": 1.4741127083068313, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.47973918914795, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8714224696159363, "num_tokens": 442154675.0, "step": 11588 }, { "epoch": 1.4742399185854218, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.419400215148926, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.875503659248352, "num_tokens": 442189449.0, "step": 11589 }, { "epoch": 1.4743671288640123, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.465725898742676, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8714267015457153, "num_tokens": 442222844.0, "step": 11590 }, { "epoch": 1.4744943391426026, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.374485969543457, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.87611323595047, "num_tokens": 442258599.0, "step": 11591 }, { "epoch": 1.4746215494211932, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.406017303466797, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8761041760444641, "num_tokens": 442294775.0, "step": 11592 }, { "epoch": 1.4747487596997837, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.43563461303711, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8802085518836975, "num_tokens": 442330313.0, "step": 11593 }, { "epoch": 1.4748759699783742, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.448850631713867, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8678156137466431, "num_tokens": 442364988.0, "step": 11594 }, { "epoch": 1.4750031802569648, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.390007019042969, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8710104823112488, "num_tokens": 442406927.0, "step": 11595 }, { "epoch": 1.4751303905355553, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.40792465209961, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8727107048034668, "num_tokens": 442442092.0, "step": 11596 }, { "epoch": 1.4752576008141458, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.451231002807617, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8763284683227539, "num_tokens": 442483013.0, "step": 11597 }, { "epoch": 1.4753848110927363, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.410643577575684, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8756917715072632, "num_tokens": 442525052.0, "step": 11598 }, { "epoch": 1.4755120213713269, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.41015625, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8604968190193176, "num_tokens": 442562543.0, "step": 11599 }, { "epoch": 1.4756392316499174, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.438091278076172, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8748559951782227, "num_tokens": 442604976.0, "step": 11600 }, { "epoch": 1.4757664419285077, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.367345809936523, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.881213903427124, "num_tokens": 442642271.0, "step": 11601 }, { "epoch": 1.4758936522070982, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.432537078857422, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.867235004901886, "num_tokens": 442680508.0, "step": 11602 }, { "epoch": 1.4760208624856888, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.444429397583008, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8584032654762268, "num_tokens": 442718391.0, "step": 11603 }, { "epoch": 1.4761480727642793, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.397706031799316, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8803285360336304, "num_tokens": 442757688.0, "step": 11604 }, { "epoch": 1.4762752830428698, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.441381454467773, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8829057216644287, "num_tokens": 442796704.0, "step": 11605 }, { "epoch": 1.4764024933214603, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.424595832824707, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.879328727722168, "num_tokens": 442838791.0, "step": 11606 }, { "epoch": 1.4765297036000509, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.411779403686523, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8703404664993286, "num_tokens": 442882368.0, "step": 11607 }, { "epoch": 1.4766569138786414, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.44898509979248, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8691426515579224, "num_tokens": 442923768.0, "step": 11608 }, { "epoch": 1.476784124157232, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.424261093139648, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8830229043960571, "num_tokens": 442955656.0, "step": 11609 }, { "epoch": 1.4769113344358225, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.440594673156738, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8602666854858398, "num_tokens": 442997323.0, "step": 11610 }, { "epoch": 1.477038544714413, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.424966812133789, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8561566472053528, "num_tokens": 443033454.0, "step": 11611 }, { "epoch": 1.4771657549930035, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.410097122192383, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8695119619369507, "num_tokens": 443077902.0, "step": 11612 }, { "epoch": 1.477292965271594, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.432483673095703, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8610261082649231, "num_tokens": 443116515.0, "step": 11613 }, { "epoch": 1.4774201755501846, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.422380447387695, "learning_rate": 1e-06, "loss": 0.3229, "mean_token_accuracy": 0.8948315382003784, "num_tokens": 443154124.0, "step": 11614 }, { "epoch": 1.477547385828775, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.416462898254395, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8774192333221436, "num_tokens": 443195852.0, "step": 11615 }, { "epoch": 1.4776745961073654, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.471552848815918, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.863593578338623, "num_tokens": 443229850.0, "step": 11616 }, { "epoch": 1.477801806385956, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.481703758239746, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8614858984947205, "num_tokens": 443274481.0, "step": 11617 }, { "epoch": 1.4779290166645465, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.44813060760498, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8756555318832397, "num_tokens": 443311737.0, "step": 11618 }, { "epoch": 1.478056226943137, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.409372329711914, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8615570664405823, "num_tokens": 443345310.0, "step": 11619 }, { "epoch": 1.4781834372217275, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.465020179748535, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8680204153060913, "num_tokens": 443392869.0, "step": 11620 }, { "epoch": 1.478310647500318, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.483805656433105, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8681254982948303, "num_tokens": 443432842.0, "step": 11621 }, { "epoch": 1.4784378577789086, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.419525146484375, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8710548877716064, "num_tokens": 443465954.0, "step": 11622 }, { "epoch": 1.478565068057499, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.48322868347168, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8609296083450317, "num_tokens": 443508506.0, "step": 11623 }, { "epoch": 1.4786922783360896, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.447461128234863, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8687776327133179, "num_tokens": 443543575.0, "step": 11624 }, { "epoch": 1.47881948861468, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.436029434204102, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.871943473815918, "num_tokens": 443578504.0, "step": 11625 }, { "epoch": 1.4789466988932705, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.46262264251709, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8823928833007812, "num_tokens": 443614647.0, "step": 11626 }, { "epoch": 1.479073909171861, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 12.412185668945312, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8650842905044556, "num_tokens": 443649244.0, "step": 11627 }, { "epoch": 1.4792011194504515, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.474492073059082, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8654516935348511, "num_tokens": 443688374.0, "step": 11628 }, { "epoch": 1.479328329729042, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.463444709777832, "learning_rate": 1e-06, "loss": 0.5023, "mean_token_accuracy": 0.8400216698646545, "num_tokens": 443730929.0, "step": 11629 }, { "epoch": 1.4794555400076326, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.487593650817871, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.86490398645401, "num_tokens": 443763067.0, "step": 11630 }, { "epoch": 1.479582750286223, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.463817596435547, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8581124544143677, "num_tokens": 443801273.0, "step": 11631 }, { "epoch": 1.4797099605648136, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.426473617553711, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8562566041946411, "num_tokens": 443840394.0, "step": 11632 }, { "epoch": 1.4798371708434042, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.484753608703613, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8655650615692139, "num_tokens": 443877486.0, "step": 11633 }, { "epoch": 1.4799643811219947, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.381437301635742, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8686131238937378, "num_tokens": 443913763.0, "step": 11634 }, { "epoch": 1.4800915914005852, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.45285415649414, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8767337799072266, "num_tokens": 443946840.0, "step": 11635 }, { "epoch": 1.4802188016791757, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.443791389465332, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8539397716522217, "num_tokens": 443992439.0, "step": 11636 }, { "epoch": 1.4803460119577663, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.410975456237793, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8793415427207947, "num_tokens": 444029348.0, "step": 11637 }, { "epoch": 1.4804732222363568, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.402557373046875, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8775257468223572, "num_tokens": 444068386.0, "step": 11638 }, { "epoch": 1.4806004325149473, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.433527946472168, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8581207394599915, "num_tokens": 444110307.0, "step": 11639 }, { "epoch": 1.4807276427935376, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.394183158874512, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8699712753295898, "num_tokens": 444153919.0, "step": 11640 }, { "epoch": 1.4808548530721282, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.414167404174805, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8710569739341736, "num_tokens": 444190575.0, "step": 11641 }, { "epoch": 1.4809820633507187, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.438726425170898, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8627088665962219, "num_tokens": 444226108.0, "step": 11642 }, { "epoch": 1.4811092736293092, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.424786567687988, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8661728501319885, "num_tokens": 444268162.0, "step": 11643 }, { "epoch": 1.4812364839078997, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.416078567504883, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8551453351974487, "num_tokens": 444309273.0, "step": 11644 }, { "epoch": 1.4813636941864903, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.469061851501465, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8733278512954712, "num_tokens": 444338697.0, "step": 11645 }, { "epoch": 1.4814909044650808, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.443899154663086, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.878553032875061, "num_tokens": 444377954.0, "step": 11646 }, { "epoch": 1.4816181147436713, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.451667785644531, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8695871829986572, "num_tokens": 444416996.0, "step": 11647 }, { "epoch": 1.4817453250222619, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.373759269714355, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8754558563232422, "num_tokens": 444455550.0, "step": 11648 }, { "epoch": 1.4818725353008524, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.476701736450195, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8651300668716431, "num_tokens": 444493979.0, "step": 11649 }, { "epoch": 1.4819997455794427, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.374773025512695, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8682379722595215, "num_tokens": 444540622.0, "step": 11650 }, { "epoch": 1.4821269558580332, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.470052719116211, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8661956191062927, "num_tokens": 444577772.0, "step": 11651 }, { "epoch": 1.4822541661366238, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.476630210876465, "learning_rate": 1e-06, "loss": 0.4709, "mean_token_accuracy": 0.8558405637741089, "num_tokens": 444615523.0, "step": 11652 }, { "epoch": 1.4823813764152143, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.43602180480957, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8674247860908508, "num_tokens": 444659785.0, "step": 11653 }, { "epoch": 1.4825085866938048, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.441600799560547, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8652921915054321, "num_tokens": 444700806.0, "step": 11654 }, { "epoch": 1.4826357969723953, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.441490173339844, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8699064254760742, "num_tokens": 444734320.0, "step": 11655 }, { "epoch": 1.4827630072509859, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.468306541442871, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8631377220153809, "num_tokens": 444770215.0, "step": 11656 }, { "epoch": 1.4828902175295764, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.431407928466797, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8676275014877319, "num_tokens": 444805432.0, "step": 11657 }, { "epoch": 1.483017427808167, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.477123260498047, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8687455654144287, "num_tokens": 444832398.0, "step": 11658 }, { "epoch": 1.4831446380867574, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.378764152526855, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8722137212753296, "num_tokens": 444876128.0, "step": 11659 }, { "epoch": 1.483271848365348, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.418534278869629, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8646580576896667, "num_tokens": 444915188.0, "step": 11660 }, { "epoch": 1.4833990586439385, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.43017292022705, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.873404860496521, "num_tokens": 444953528.0, "step": 11661 }, { "epoch": 1.483526268922529, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.42213249206543, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8619429469108582, "num_tokens": 444996469.0, "step": 11662 }, { "epoch": 1.4836534792011196, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.38381290435791, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8860793113708496, "num_tokens": 445034298.0, "step": 11663 }, { "epoch": 1.48378068947971, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.476520538330078, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8678380846977234, "num_tokens": 445079010.0, "step": 11664 }, { "epoch": 1.4839078997583004, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.435996055603027, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8828051090240479, "num_tokens": 445116009.0, "step": 11665 }, { "epoch": 1.484035110036891, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.432573318481445, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8813836574554443, "num_tokens": 445156313.0, "step": 11666 }, { "epoch": 1.4841623203154815, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.481860160827637, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8701788187026978, "num_tokens": 445187424.0, "step": 11667 }, { "epoch": 1.484289530594072, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.486420631408691, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8695025444030762, "num_tokens": 445233448.0, "step": 11668 }, { "epoch": 1.4844167408726625, "ewc_loss": 0.02880859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.42540454864502, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8803693056106567, "num_tokens": 445273663.0, "step": 11669 }, { "epoch": 1.484543951151253, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.460461616516113, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8753334879875183, "num_tokens": 445313790.0, "step": 11670 }, { "epoch": 1.4846711614298436, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.459598541259766, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8789349794387817, "num_tokens": 445351818.0, "step": 11671 }, { "epoch": 1.484798371708434, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.45645809173584, "learning_rate": 1e-06, "loss": 0.4896, "mean_token_accuracy": 0.844096302986145, "num_tokens": 445391763.0, "step": 11672 }, { "epoch": 1.4849255819870246, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.429184913635254, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8625682592391968, "num_tokens": 445426723.0, "step": 11673 }, { "epoch": 1.485052792265615, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.4216890335083, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.874033510684967, "num_tokens": 445465635.0, "step": 11674 }, { "epoch": 1.4851800025442055, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.423093795776367, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8700541257858276, "num_tokens": 445502819.0, "step": 11675 }, { "epoch": 1.485307212822796, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.44693374633789, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.878717303276062, "num_tokens": 445541861.0, "step": 11676 }, { "epoch": 1.4854344231013865, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.441500663757324, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8844224214553833, "num_tokens": 445578058.0, "step": 11677 }, { "epoch": 1.485561633379977, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.472213745117188, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.854637861251831, "num_tokens": 445612896.0, "step": 11678 }, { "epoch": 1.4856888436585676, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.457913398742676, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8622561097145081, "num_tokens": 445647849.0, "step": 11679 }, { "epoch": 1.485816053937158, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.41440200805664, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8540531992912292, "num_tokens": 445689199.0, "step": 11680 }, { "epoch": 1.4859432642157486, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.464018821716309, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8676559329032898, "num_tokens": 445727869.0, "step": 11681 }, { "epoch": 1.4860704744943392, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.424056053161621, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.872138261795044, "num_tokens": 445770105.0, "step": 11682 }, { "epoch": 1.4861976847729297, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.471485137939453, "learning_rate": 1e-06, "loss": 0.4719, "mean_token_accuracy": 0.8489357829093933, "num_tokens": 445812950.0, "step": 11683 }, { "epoch": 1.4863248950515202, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.430985450744629, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8668981790542603, "num_tokens": 445850811.0, "step": 11684 }, { "epoch": 1.4864521053301107, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.437323570251465, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8714927434921265, "num_tokens": 445891967.0, "step": 11685 }, { "epoch": 1.4865793156087013, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.422266006469727, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8708618879318237, "num_tokens": 445927936.0, "step": 11686 }, { "epoch": 1.4867065258872918, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.455636024475098, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.881439208984375, "num_tokens": 445962481.0, "step": 11687 }, { "epoch": 1.4868337361658823, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.435208320617676, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8819727897644043, "num_tokens": 446001445.0, "step": 11688 }, { "epoch": 1.4869609464444726, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.394712448120117, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8883593082427979, "num_tokens": 446037904.0, "step": 11689 }, { "epoch": 1.4870881567230632, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.508238792419434, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8778923749923706, "num_tokens": 446077583.0, "step": 11690 }, { "epoch": 1.4872153670016537, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.459001541137695, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8682071566581726, "num_tokens": 446110123.0, "step": 11691 }, { "epoch": 1.4873425772802442, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.445435523986816, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8669167160987854, "num_tokens": 446149832.0, "step": 11692 }, { "epoch": 1.4874697875588347, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.394002914428711, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8759106397628784, "num_tokens": 446184211.0, "step": 11693 }, { "epoch": 1.4875969978374253, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.44421672821045, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8630831241607666, "num_tokens": 446229620.0, "step": 11694 }, { "epoch": 1.4877242081160158, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.445323944091797, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8709220886230469, "num_tokens": 446267134.0, "step": 11695 }, { "epoch": 1.4878514183946063, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.403294563293457, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8576964139938354, "num_tokens": 446303001.0, "step": 11696 }, { "epoch": 1.4879786286731969, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.498873710632324, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8682852983474731, "num_tokens": 446334286.0, "step": 11697 }, { "epoch": 1.4881058389517874, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.426074028015137, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.886012852191925, "num_tokens": 446369547.0, "step": 11698 }, { "epoch": 1.4882330492303777, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.510939598083496, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8539674282073975, "num_tokens": 446408179.0, "step": 11699 }, { "epoch": 1.4883602595089682, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.434067726135254, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8712661266326904, "num_tokens": 446448533.0, "step": 11700 }, { "epoch": 1.4884874697875587, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.46818733215332, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.866308331489563, "num_tokens": 446489445.0, "step": 11701 }, { "epoch": 1.4886146800661493, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.420170783996582, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8831048011779785, "num_tokens": 446526530.0, "step": 11702 }, { "epoch": 1.4887418903447398, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.502185821533203, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8701738715171814, "num_tokens": 446558315.0, "step": 11703 }, { "epoch": 1.4888691006233303, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.385995864868164, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.87270188331604, "num_tokens": 446598548.0, "step": 11704 }, { "epoch": 1.4889963109019209, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.515612602233887, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.865429162979126, "num_tokens": 446632111.0, "step": 11705 }, { "epoch": 1.4891235211805114, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.421873092651367, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8767921328544617, "num_tokens": 446669296.0, "step": 11706 }, { "epoch": 1.489250731459102, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.46971321105957, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8713638782501221, "num_tokens": 446707891.0, "step": 11707 }, { "epoch": 1.4893779417376924, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.398833274841309, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.858191967010498, "num_tokens": 446749455.0, "step": 11708 }, { "epoch": 1.489505152016283, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.492928504943848, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8814572095870972, "num_tokens": 446787001.0, "step": 11709 }, { "epoch": 1.4896323622948735, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.451614379882812, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8691717386245728, "num_tokens": 446827262.0, "step": 11710 }, { "epoch": 1.489759572573464, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.454038619995117, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8842931389808655, "num_tokens": 446860514.0, "step": 11711 }, { "epoch": 1.4898867828520546, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.46005916595459, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8604954481124878, "num_tokens": 446902605.0, "step": 11712 }, { "epoch": 1.490013993130645, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.46786117553711, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8761197328567505, "num_tokens": 446938638.0, "step": 11713 }, { "epoch": 1.4901412034092354, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.461913108825684, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8749112486839294, "num_tokens": 446979286.0, "step": 11714 }, { "epoch": 1.490268413687826, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.4720458984375, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8544689416885376, "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": 12.42916488647461, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8713265061378479, "num_tokens": 447056733.0, "step": 11716 }, { "epoch": 1.490522834245007, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.468233108520508, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8649988174438477, "num_tokens": 447094493.0, "step": 11717 }, { "epoch": 1.4906500445235975, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.483071327209473, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8845451474189758, "num_tokens": 447134039.0, "step": 11718 }, { "epoch": 1.490777254802188, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.453095436096191, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8781155347824097, "num_tokens": 447168320.0, "step": 11719 }, { "epoch": 1.4909044650807786, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.521997451782227, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8540345430374146, "num_tokens": 447205640.0, "step": 11720 }, { "epoch": 1.491031675359369, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.479842185974121, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8833975791931152, "num_tokens": 447239831.0, "step": 11721 }, { "epoch": 1.4911588856379596, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.44536304473877, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.854399561882019, "num_tokens": 447281100.0, "step": 11722 }, { "epoch": 1.49128609591655, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.481184005737305, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.850052535533905, "num_tokens": 447316137.0, "step": 11723 }, { "epoch": 1.4914133061951405, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.447490692138672, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8623934984207153, "num_tokens": 447357922.0, "step": 11724 }, { "epoch": 1.491540516473731, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.462246894836426, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.866206169128418, "num_tokens": 447394905.0, "step": 11725 }, { "epoch": 1.4916677267523215, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.453728675842285, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8798548579216003, "num_tokens": 447435379.0, "step": 11726 }, { "epoch": 1.491794937030912, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.486016273498535, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8609819412231445, "num_tokens": 447473058.0, "step": 11727 }, { "epoch": 1.4919221473095026, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.422898292541504, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8785617351531982, "num_tokens": 447510966.0, "step": 11728 }, { "epoch": 1.492049357588093, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.476825714111328, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8725696802139282, "num_tokens": 447554747.0, "step": 11729 }, { "epoch": 1.4921765678666836, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.462098121643066, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8718424439430237, "num_tokens": 447594426.0, "step": 11730 }, { "epoch": 1.4923037781452742, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.496541023254395, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8622156381607056, "num_tokens": 447634786.0, "step": 11731 }, { "epoch": 1.4924309884238647, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.524524688720703, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8672609925270081, "num_tokens": 447668682.0, "step": 11732 }, { "epoch": 1.4925581987024552, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.446869850158691, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8615190386772156, "num_tokens": 447705566.0, "step": 11733 }, { "epoch": 1.4926854089810457, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.44141960144043, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8667449951171875, "num_tokens": 447746385.0, "step": 11734 }, { "epoch": 1.4928126192596363, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.494359016418457, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8735008835792542, "num_tokens": 447790483.0, "step": 11735 }, { "epoch": 1.4929398295382268, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.50351333618164, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8675329089164734, "num_tokens": 447826872.0, "step": 11736 }, { "epoch": 1.4930670398168173, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.452653884887695, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8737047910690308, "num_tokens": 447864071.0, "step": 11737 }, { "epoch": 1.4931942500954076, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.456326484680176, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8515949249267578, "num_tokens": 447903320.0, "step": 11738 }, { "epoch": 1.4933214603739982, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.511796951293945, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8829430937767029, "num_tokens": 447940253.0, "step": 11739 }, { "epoch": 1.4934486706525887, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.413710594177246, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8746116161346436, "num_tokens": 447978277.0, "step": 11740 }, { "epoch": 1.4935758809311792, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.446739196777344, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8579874634742737, "num_tokens": 448018798.0, "step": 11741 }, { "epoch": 1.4937030912097697, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.541439056396484, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8593236207962036, "num_tokens": 448057469.0, "step": 11742 }, { "epoch": 1.4938303014883603, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.502946853637695, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8572636246681213, "num_tokens": 448089032.0, "step": 11743 }, { "epoch": 1.4939575117669508, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.539285659790039, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8648065328598022, "num_tokens": 448122438.0, "step": 11744 }, { "epoch": 1.4940847220455413, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.449857711791992, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8677552342414856, "num_tokens": 448166844.0, "step": 11745 }, { "epoch": 1.4942119323241319, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.504010200500488, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8601491451263428, "num_tokens": 448197766.0, "step": 11746 }, { "epoch": 1.4943391426027224, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.496075630187988, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8562904596328735, "num_tokens": 448240850.0, "step": 11747 }, { "epoch": 1.4944663528813127, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.461041450500488, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8781538009643555, "num_tokens": 448274187.0, "step": 11748 }, { "epoch": 1.4945935631599032, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.462546348571777, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8749340772628784, "num_tokens": 448311069.0, "step": 11749 }, { "epoch": 1.4947207734384937, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.48083782196045, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8688551187515259, "num_tokens": 448351536.0, "step": 11750 }, { "epoch": 1.4948479837170843, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.483306884765625, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.86604905128479, "num_tokens": 448385108.0, "step": 11751 }, { "epoch": 1.4949751939956748, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.479551315307617, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8798247575759888, "num_tokens": 448423272.0, "step": 11752 }, { "epoch": 1.4951024042742653, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.451475143432617, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8744882345199585, "num_tokens": 448462904.0, "step": 11753 }, { "epoch": 1.4952296145528559, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.494654655456543, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8653096556663513, "num_tokens": 448503257.0, "step": 11754 }, { "epoch": 1.4953568248314464, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.531630516052246, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8749275207519531, "num_tokens": 448535299.0, "step": 11755 }, { "epoch": 1.495484035110037, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.485977172851562, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8756789565086365, "num_tokens": 448568878.0, "step": 11756 }, { "epoch": 1.4956112453886274, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.469785690307617, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8879408836364746, "num_tokens": 448604959.0, "step": 11757 }, { "epoch": 1.495738455667218, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.477048873901367, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8807579278945923, "num_tokens": 448638487.0, "step": 11758 }, { "epoch": 1.4958656659458085, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.483954429626465, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8751097321510315, "num_tokens": 448677212.0, "step": 11759 }, { "epoch": 1.495992876224399, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.482955932617188, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.864812970161438, "num_tokens": 448719654.0, "step": 11760 }, { "epoch": 1.4961200865029896, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.48017692565918, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8729516267776489, "num_tokens": 448760783.0, "step": 11761 }, { "epoch": 1.49624729678158, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.504375457763672, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8583976626396179, "num_tokens": 448802843.0, "step": 11762 }, { "epoch": 1.4963745070601704, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.508500099182129, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8653996586799622, "num_tokens": 448838430.0, "step": 11763 }, { "epoch": 1.496501717338761, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.468450546264648, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8607125282287598, "num_tokens": 448880450.0, "step": 11764 }, { "epoch": 1.4966289276173514, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.490099906921387, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8722413182258606, "num_tokens": 448919352.0, "step": 11765 }, { "epoch": 1.496756137895942, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.467965126037598, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8445271253585815, "num_tokens": 448965402.0, "step": 11766 }, { "epoch": 1.4968833481745325, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.483501434326172, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8811757564544678, "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": 12.460412979125977, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8564857840538025, "num_tokens": 449036555.0, "step": 11768 }, { "epoch": 1.4971377687317136, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.488541603088379, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.880518913269043, "num_tokens": 449070545.0, "step": 11769 }, { "epoch": 1.497264979010304, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.440842628479004, "learning_rate": 1e-06, "loss": 0.4748, "mean_token_accuracy": 0.8536648750305176, "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": 12.462545394897461, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8667280077934265, "num_tokens": 449145333.0, "step": 11771 }, { "epoch": 1.497519399567485, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.484556198120117, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8627817630767822, "num_tokens": 449181292.0, "step": 11772 }, { "epoch": 1.4976466098460754, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.473410606384277, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8725619316101074, "num_tokens": 449213940.0, "step": 11773 }, { "epoch": 1.497773820124666, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.415282249450684, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8793642520904541, "num_tokens": 449252487.0, "step": 11774 }, { "epoch": 1.4979010304032565, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.462830543518066, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8617932796478271, "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": 12.421544075012207, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.864827036857605, "num_tokens": 449326975.0, "step": 11776 }, { "epoch": 1.4981554509604376, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.481666564941406, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8682374954223633, "num_tokens": 449366061.0, "step": 11777 }, { "epoch": 1.498282661239028, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.451534271240234, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8767353892326355, "num_tokens": 449401763.0, "step": 11778 }, { "epoch": 1.4984098715176186, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.550492286682129, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8884457349777222, "num_tokens": 449436605.0, "step": 11779 }, { "epoch": 1.4985370817962091, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.469347953796387, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.86451256275177, "num_tokens": 449474523.0, "step": 11780 }, { "epoch": 1.4986642920747997, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.511595726013184, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8813483715057373, "num_tokens": 449513301.0, "step": 11781 }, { "epoch": 1.4987915023533902, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.499667167663574, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8655471801757812, "num_tokens": 449547826.0, "step": 11782 }, { "epoch": 1.4989187126319807, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.549365043640137, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8685638904571533, "num_tokens": 449584818.0, "step": 11783 }, { "epoch": 1.4990459229105713, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.513862609863281, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8506042957305908, "num_tokens": 449625368.0, "step": 11784 }, { "epoch": 1.4991731331891618, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 12.469345092773438, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8574023842811584, "num_tokens": 449666320.0, "step": 11785 }, { "epoch": 1.4993003434677523, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.550859451293945, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8600411415100098, "num_tokens": 449705993.0, "step": 11786 }, { "epoch": 1.4994275537463426, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.46233081817627, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8646864891052246, "num_tokens": 449747088.0, "step": 11787 }, { "epoch": 1.4995547640249332, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.473862648010254, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8710365891456604, "num_tokens": 449789840.0, "step": 11788 }, { "epoch": 1.4996819743035237, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.515323638916016, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8681559562683105, "num_tokens": 449830599.0, "step": 11789 }, { "epoch": 1.4998091845821142, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.549300193786621, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8732956647872925, "num_tokens": 449864476.0, "step": 11790 }, { "epoch": 1.4999363948607047, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.546433448791504, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8531380891799927, "num_tokens": 449903791.0, "step": 11791 }, { "epoch": 1.5000636051392953, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.532313346862793, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.869275689125061, "num_tokens": 449945478.0, "step": 11792 }, { "epoch": 1.5001908154178858, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.440841674804688, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.866540253162384, "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": 12.530091285705566, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8660410642623901, "num_tokens": 450025537.0, "step": 11794 }, { "epoch": 1.5004452359750666, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.5158052444458, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8540148735046387, "num_tokens": 450068269.0, "step": 11795 }, { "epoch": 1.5005724462536572, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.454103469848633, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8841884136199951, "num_tokens": 450106234.0, "step": 11796 }, { "epoch": 1.5006996565322477, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.504602432250977, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8781360387802124, "num_tokens": 450141916.0, "step": 11797 }, { "epoch": 1.5008268668108382, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.530444145202637, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8551371097564697, "num_tokens": 450175427.0, "step": 11798 }, { "epoch": 1.5009540770894287, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.508708953857422, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8716170191764832, "num_tokens": 450211104.0, "step": 11799 }, { "epoch": 1.5010812873680193, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.547553062438965, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8609687089920044, "num_tokens": 450244844.0, "step": 11800 }, { "epoch": 1.5012084976466098, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.464031219482422, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8683549165725708, "num_tokens": 450286094.0, "step": 11801 }, { "epoch": 1.5013357079252003, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.50342082977295, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.8605191707611084, "num_tokens": 450324938.0, "step": 11802 }, { "epoch": 1.5014629182037909, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.520708084106445, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8692943453788757, "num_tokens": 450368572.0, "step": 11803 }, { "epoch": 1.5015901284823814, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.470486640930176, "learning_rate": 1e-06, "loss": 0.482, "mean_token_accuracy": 0.8479049205780029, "num_tokens": 450410926.0, "step": 11804 }, { "epoch": 1.501717338760972, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.50827407836914, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8724161386489868, "num_tokens": 450450206.0, "step": 11805 }, { "epoch": 1.5018445490395624, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.533964157104492, "learning_rate": 1e-06, "loss": 0.4786, "mean_token_accuracy": 0.8491368889808655, "num_tokens": 450488998.0, "step": 11806 }, { "epoch": 1.501971759318153, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.549383163452148, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8794058561325073, "num_tokens": 450530476.0, "step": 11807 }, { "epoch": 1.5020989695967435, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.516225814819336, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.875705361366272, "num_tokens": 450560655.0, "step": 11808 }, { "epoch": 1.502226179875334, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.51229476928711, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8712027072906494, "num_tokens": 450601816.0, "step": 11809 }, { "epoch": 1.5023533901539246, "ewc_loss": 0.029052734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.496665000915527, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8726475238800049, "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": 12.47765827178955, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8649826049804688, "num_tokens": 450672705.0, "step": 11811 }, { "epoch": 1.5026078107111056, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.530821800231934, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8561024069786072, "num_tokens": 450705636.0, "step": 11812 }, { "epoch": 1.502735020989696, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.516472816467285, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8604453206062317, "num_tokens": 450736165.0, "step": 11813 }, { "epoch": 1.5028622312682864, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.514056205749512, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.870346188545227, "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": 12.457148551940918, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8726588487625122, "num_tokens": 450811994.0, "step": 11815 }, { "epoch": 1.5031166518254675, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.439852714538574, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8692464828491211, "num_tokens": 450853688.0, "step": 11816 }, { "epoch": 1.503243862104058, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.526274681091309, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.869001567363739, "num_tokens": 450892100.0, "step": 11817 }, { "epoch": 1.5033710723826486, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.475112915039062, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8853119015693665, "num_tokens": 450936383.0, "step": 11818 }, { "epoch": 1.503498282661239, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.542424201965332, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.876121997833252, "num_tokens": 450965525.0, "step": 11819 }, { "epoch": 1.5036254929398294, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.467415809631348, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8565086722373962, "num_tokens": 451005874.0, "step": 11820 }, { "epoch": 1.50375270321842, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.501160621643066, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8495379686355591, "num_tokens": 451040668.0, "step": 11821 }, { "epoch": 1.5038799134970104, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.488482475280762, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8658247590065002, "num_tokens": 451082747.0, "step": 11822 }, { "epoch": 1.504007123775601, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.50059986114502, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8751718997955322, "num_tokens": 451120344.0, "step": 11823 }, { "epoch": 1.5041343340541915, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.516528129577637, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8629571199417114, "num_tokens": 451156564.0, "step": 11824 }, { "epoch": 1.504261544332782, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.457613945007324, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8639694452285767, "num_tokens": 451197252.0, "step": 11825 }, { "epoch": 1.5043887546113726, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.534337997436523, "learning_rate": 1e-06, "loss": 0.4652, "mean_token_accuracy": 0.8575350046157837, "num_tokens": 451236150.0, "step": 11826 }, { "epoch": 1.504515964889963, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.475119590759277, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8643239736557007, "num_tokens": 451275751.0, "step": 11827 }, { "epoch": 1.5046431751685536, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.450610160827637, "learning_rate": 1e-06, "loss": 0.3427, "mean_token_accuracy": 0.8894906640052795, "num_tokens": 451316778.0, "step": 11828 }, { "epoch": 1.5047703854471441, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.537978172302246, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8671396970748901, "num_tokens": 451359921.0, "step": 11829 }, { "epoch": 1.5048975957257347, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.459518432617188, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8748040795326233, "num_tokens": 451395616.0, "step": 11830 }, { "epoch": 1.5050248060043252, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.550748825073242, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8676652908325195, "num_tokens": 451427766.0, "step": 11831 }, { "epoch": 1.5051520162829157, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.436997413635254, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8770059943199158, "num_tokens": 451463073.0, "step": 11832 }, { "epoch": 1.5052792265615063, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.50485610961914, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8877366781234741, "num_tokens": 451503039.0, "step": 11833 }, { "epoch": 1.5054064368400968, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.544478416442871, "learning_rate": 1e-06, "loss": 0.4852, "mean_token_accuracy": 0.8516059517860413, "num_tokens": 451541599.0, "step": 11834 }, { "epoch": 1.5055336471186873, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.506757736206055, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8541021943092346, "num_tokens": 451579278.0, "step": 11835 }, { "epoch": 1.5056608573972778, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.506035804748535, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8718156218528748, "num_tokens": 451618591.0, "step": 11836 }, { "epoch": 1.5057880676758684, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.52400016784668, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8688672780990601, "num_tokens": 451654453.0, "step": 11837 }, { "epoch": 1.5059152779544587, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.572514533996582, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8673373460769653, "num_tokens": 451688302.0, "step": 11838 }, { "epoch": 1.5060424882330492, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.494855880737305, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8741580247879028, "num_tokens": 451727301.0, "step": 11839 }, { "epoch": 1.5061696985116397, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.560709953308105, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8547557592391968, "num_tokens": 451770471.0, "step": 11840 }, { "epoch": 1.5062969087902303, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.522061347961426, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8570156693458557, "num_tokens": 451809413.0, "step": 11841 }, { "epoch": 1.5064241190688208, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.552469253540039, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8660920858383179, "num_tokens": 451850177.0, "step": 11842 }, { "epoch": 1.5065513293474113, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.52756404876709, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8560084700584412, "num_tokens": 451887977.0, "step": 11843 }, { "epoch": 1.5066785396260016, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.529916763305664, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8795439004898071, "num_tokens": 451927403.0, "step": 11844 }, { "epoch": 1.5068057499045922, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.536735534667969, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8678399324417114, "num_tokens": 451967350.0, "step": 11845 }, { "epoch": 1.5069329601831827, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.510361671447754, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8678119778633118, "num_tokens": 452008664.0, "step": 11846 }, { "epoch": 1.5070601704617732, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.51709270477295, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.874953031539917, "num_tokens": 452047303.0, "step": 11847 }, { "epoch": 1.5071873807403637, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.560443878173828, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8700281381607056, "num_tokens": 452087359.0, "step": 11848 }, { "epoch": 1.5073145910189543, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.498307228088379, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.854155421257019, "num_tokens": 452129426.0, "step": 11849 }, { "epoch": 1.5074418012975448, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.519791603088379, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8731226921081543, "num_tokens": 452169134.0, "step": 11850 }, { "epoch": 1.5075690115761353, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.565982818603516, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8592392802238464, "num_tokens": 452209962.0, "step": 11851 }, { "epoch": 1.5076962218547258, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.504915237426758, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8715329766273499, "num_tokens": 452245978.0, "step": 11852 }, { "epoch": 1.5078234321333164, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.52759075164795, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8785767555236816, "num_tokens": 452281013.0, "step": 11853 }, { "epoch": 1.507950642411907, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.544014930725098, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8823356628417969, "num_tokens": 452317174.0, "step": 11854 }, { "epoch": 1.5080778526904974, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.513199806213379, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8647786378860474, "num_tokens": 452362974.0, "step": 11855 }, { "epoch": 1.508205062969088, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.5097074508667, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8793395161628723, "num_tokens": 452400701.0, "step": 11856 }, { "epoch": 1.5083322732476785, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.487845420837402, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8782511949539185, "num_tokens": 452438541.0, "step": 11857 }, { "epoch": 1.508459483526269, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.491842269897461, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8628127574920654, "num_tokens": 452476918.0, "step": 11858 }, { "epoch": 1.5085866938048595, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.517041206359863, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8723803758621216, "num_tokens": 452509231.0, "step": 11859 }, { "epoch": 1.50871390408345, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.486811637878418, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8631879687309265, "num_tokens": 452547079.0, "step": 11860 }, { "epoch": 1.5088411143620406, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53101634979248, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8794382214546204, "num_tokens": 452580215.0, "step": 11861 }, { "epoch": 1.508968324640631, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.50120735168457, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8752438426017761, "num_tokens": 452615778.0, "step": 11862 }, { "epoch": 1.5090955349192214, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.475707054138184, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8713065981864929, "num_tokens": 452652184.0, "step": 11863 }, { "epoch": 1.509222745197812, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.511663436889648, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8541982769966125, "num_tokens": 452691345.0, "step": 11864 }, { "epoch": 1.5093499554764025, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.554980278015137, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8694167137145996, "num_tokens": 452729982.0, "step": 11865 }, { "epoch": 1.509477165754993, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.431146621704102, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8682183027267456, "num_tokens": 452771272.0, "step": 11866 }, { "epoch": 1.5096043760335836, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.542743682861328, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8648077249526978, "num_tokens": 452812231.0, "step": 11867 }, { "epoch": 1.509731586312174, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.474478721618652, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8572700023651123, "num_tokens": 452858406.0, "step": 11868 }, { "epoch": 1.5098587965907644, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.428116798400879, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8601271510124207, "num_tokens": 452898793.0, "step": 11869 }, { "epoch": 1.509986006869355, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.550819396972656, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8730313181877136, "num_tokens": 452937929.0, "step": 11870 }, { "epoch": 1.5101132171479454, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.52649211883545, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8714958429336548, "num_tokens": 452975436.0, "step": 11871 }, { "epoch": 1.510240427426536, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.513862609863281, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8688068389892578, "num_tokens": 453016334.0, "step": 11872 }, { "epoch": 1.5103676377051265, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.528940200805664, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8722283840179443, "num_tokens": 453053630.0, "step": 11873 }, { "epoch": 1.510494847983717, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.54978084564209, "learning_rate": 1e-06, "loss": 0.4688, "mean_token_accuracy": 0.8501072525978088, "num_tokens": 453092341.0, "step": 11874 }, { "epoch": 1.5106220582623076, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.54859447479248, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8764351606369019, "num_tokens": 453126859.0, "step": 11875 }, { "epoch": 1.510749268540898, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53031063079834, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8653144240379333, "num_tokens": 453167327.0, "step": 11876 }, { "epoch": 1.5108764788194886, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.53180980682373, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8744131326675415, "num_tokens": 453205901.0, "step": 11877 }, { "epoch": 1.5110036890980791, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.492960929870605, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8671635389328003, "num_tokens": 453241856.0, "step": 11878 }, { "epoch": 1.5111308993766697, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.47036361694336, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8763132095336914, "num_tokens": 453282528.0, "step": 11879 }, { "epoch": 1.5112581096552602, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.512595176696777, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8593804836273193, "num_tokens": 453318762.0, "step": 11880 }, { "epoch": 1.5113853199338507, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.536280632019043, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8833322525024414, "num_tokens": 453352482.0, "step": 11881 }, { "epoch": 1.5115125302124413, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.502067565917969, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.868099570274353, "num_tokens": 453393692.0, "step": 11882 }, { "epoch": 1.5116397404910318, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.514540672302246, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8730190992355347, "num_tokens": 453429217.0, "step": 11883 }, { "epoch": 1.5117669507696223, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.470870971679688, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8544114828109741, "num_tokens": 453468190.0, "step": 11884 }, { "epoch": 1.5118941610482128, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.492191314697266, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8618255853652954, "num_tokens": 453509893.0, "step": 11885 }, { "epoch": 1.5120213713268034, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.493404388427734, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8715755939483643, "num_tokens": 453546972.0, "step": 11886 }, { "epoch": 1.5121485816053937, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.500665664672852, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8798725008964539, "num_tokens": 453578480.0, "step": 11887 }, { "epoch": 1.5122757918839842, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.469008445739746, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8691664934158325, "num_tokens": 453623248.0, "step": 11888 }, { "epoch": 1.5124030021625747, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.498260498046875, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8887994289398193, "num_tokens": 453660088.0, "step": 11889 }, { "epoch": 1.5125302124411653, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.501422882080078, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.870331883430481, "num_tokens": 453698931.0, "step": 11890 }, { "epoch": 1.5126574227197558, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.467205047607422, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8714863061904907, "num_tokens": 453734220.0, "step": 11891 }, { "epoch": 1.5127846329983463, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.500531196594238, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8672147989273071, "num_tokens": 453774056.0, "step": 11892 }, { "epoch": 1.5129118432769366, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.498326301574707, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.875483512878418, "num_tokens": 453806616.0, "step": 11893 }, { "epoch": 1.5130390535555271, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.498490333557129, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8772883415222168, "num_tokens": 453843838.0, "step": 11894 }, { "epoch": 1.5131662638341177, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.49092960357666, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8767997026443481, "num_tokens": 453876888.0, "step": 11895 }, { "epoch": 1.5132934741127082, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.507566452026367, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8839864730834961, "num_tokens": 453915334.0, "step": 11896 }, { "epoch": 1.5134206843912987, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.4879150390625, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8777302503585815, "num_tokens": 453950590.0, "step": 11897 }, { "epoch": 1.5135478946698893, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.503388404846191, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8726335763931274, "num_tokens": 453992590.0, "step": 11898 }, { "epoch": 1.5136751049484798, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.398327827453613, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8736589550971985, "num_tokens": 454038619.0, "step": 11899 }, { "epoch": 1.5138023152270703, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.493599891662598, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8602421283721924, "num_tokens": 454081564.0, "step": 11900 }, { "epoch": 1.5139295255056608, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.505020141601562, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8568777441978455, "num_tokens": 454125581.0, "step": 11901 }, { "epoch": 1.5140567357842514, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.522982597351074, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8773974776268005, "num_tokens": 454163961.0, "step": 11902 }, { "epoch": 1.514183946062842, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.500965118408203, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8627837896347046, "num_tokens": 454200410.0, "step": 11903 }, { "epoch": 1.5143111563414324, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.494575500488281, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.855624794960022, "num_tokens": 454237143.0, "step": 11904 }, { "epoch": 1.514438366620023, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.511799812316895, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8769232034683228, "num_tokens": 454274022.0, "step": 11905 }, { "epoch": 1.5145655768986135, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.520417213439941, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8639117479324341, "num_tokens": 454307955.0, "step": 11906 }, { "epoch": 1.514692787177204, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.509175300598145, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8763005137443542, "num_tokens": 454345449.0, "step": 11907 }, { "epoch": 1.5148199974557945, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.499007225036621, "learning_rate": 1e-06, "loss": 0.4491, "mean_token_accuracy": 0.8578343391418457, "num_tokens": 454383096.0, "step": 11908 }, { "epoch": 1.514947207734385, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.500316619873047, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8583696484565735, "num_tokens": 454420142.0, "step": 11909 }, { "epoch": 1.5150744180129756, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.516158103942871, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8676115870475769, "num_tokens": 454454991.0, "step": 11910 }, { "epoch": 1.515201628291566, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.561026573181152, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8656013011932373, "num_tokens": 454490161.0, "step": 11911 }, { "epoch": 1.5153288385701564, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.509154319763184, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8601043820381165, "num_tokens": 454534767.0, "step": 11912 }, { "epoch": 1.515456048848747, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.491579055786133, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8780926465988159, "num_tokens": 454573246.0, "step": 11913 }, { "epoch": 1.5155832591273375, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.510331153869629, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8820968866348267, "num_tokens": 454607626.0, "step": 11914 }, { "epoch": 1.515710469405928, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.47350025177002, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8620866537094116, "num_tokens": 454639840.0, "step": 11915 }, { "epoch": 1.5158376796845185, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.48737907409668, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8598676919937134, "num_tokens": 454681445.0, "step": 11916 }, { "epoch": 1.515964889963109, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.503815650939941, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8610695600509644, "num_tokens": 454726191.0, "step": 11917 }, { "epoch": 1.5160921002416994, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.514527320861816, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8754333257675171, "num_tokens": 454765112.0, "step": 11918 }, { "epoch": 1.51621931052029, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.472230911254883, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8663719892501831, "num_tokens": 454805254.0, "step": 11919 }, { "epoch": 1.5163465207988804, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.560522079467773, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8615922927856445, "num_tokens": 454841344.0, "step": 11920 }, { "epoch": 1.516473731077471, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.455161094665527, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8817569017410278, "num_tokens": 454879986.0, "step": 11921 }, { "epoch": 1.5166009413560615, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.608053207397461, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8601403832435608, "num_tokens": 454917731.0, "step": 11922 }, { "epoch": 1.516728151634652, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.52460765838623, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8743687272071838, "num_tokens": 454952837.0, "step": 11923 }, { "epoch": 1.5168553619132426, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.498327255249023, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8713253736495972, "num_tokens": 454992751.0, "step": 11924 }, { "epoch": 1.516982572191833, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.505470275878906, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8679081797599792, "num_tokens": 455029829.0, "step": 11925 }, { "epoch": 1.5171097824704236, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.55337142944336, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8687022924423218, "num_tokens": 455072167.0, "step": 11926 }, { "epoch": 1.5172369927490141, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.5136079788208, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8759298324584961, "num_tokens": 455108433.0, "step": 11927 }, { "epoch": 1.5173642030276047, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.472835540771484, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8726891279220581, "num_tokens": 455148943.0, "step": 11928 }, { "epoch": 1.5174914133061952, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.567699432373047, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8639618158340454, "num_tokens": 455185099.0, "step": 11929 }, { "epoch": 1.5176186235847857, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.467329025268555, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8823885917663574, "num_tokens": 455225322.0, "step": 11930 }, { "epoch": 1.5177458338633762, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.537687301635742, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.866400957107544, "num_tokens": 455262830.0, "step": 11931 }, { "epoch": 1.5178730441419668, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.493337631225586, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8755972981452942, "num_tokens": 455300825.0, "step": 11932 }, { "epoch": 1.5180002544205573, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.54270076751709, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.866117000579834, "num_tokens": 455336209.0, "step": 11933 }, { "epoch": 1.5181274646991478, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.511713981628418, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8685348629951477, "num_tokens": 455374909.0, "step": 11934 }, { "epoch": 1.5182546749777384, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.493724822998047, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.881203293800354, "num_tokens": 455417179.0, "step": 11935 }, { "epoch": 1.5183818852563287, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.508166313171387, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.855461597442627, "num_tokens": 455457923.0, "step": 11936 }, { "epoch": 1.5185090955349192, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.51080322265625, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8603320717811584, "num_tokens": 455492932.0, "step": 11937 }, { "epoch": 1.5186363058135097, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.564046859741211, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8594926595687866, "num_tokens": 455540392.0, "step": 11938 }, { "epoch": 1.5187635160921003, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.483485221862793, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8644787073135376, "num_tokens": 455574461.0, "step": 11939 }, { "epoch": 1.5188907263706908, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.57809829711914, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8649897575378418, "num_tokens": 455608315.0, "step": 11940 }, { "epoch": 1.5190179366492813, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.461403846740723, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.872805118560791, "num_tokens": 455654927.0, "step": 11941 }, { "epoch": 1.5191451469278716, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.554292678833008, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8557542562484741, "num_tokens": 455695745.0, "step": 11942 }, { "epoch": 1.5192723572064621, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.520241737365723, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8692458868026733, "num_tokens": 455734652.0, "step": 11943 }, { "epoch": 1.5193995674850527, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.48525333404541, "learning_rate": 1e-06, "loss": 0.4729, "mean_token_accuracy": 0.8461970090866089, "num_tokens": 455776063.0, "step": 11944 }, { "epoch": 1.5195267777636432, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.499382019042969, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.863492488861084, "num_tokens": 455820595.0, "step": 11945 }, { "epoch": 1.5196539880422337, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.484189987182617, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8738831281661987, "num_tokens": 455859553.0, "step": 11946 }, { "epoch": 1.5197811983208243, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.528461456298828, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8690721988677979, "num_tokens": 455892676.0, "step": 11947 }, { "epoch": 1.5199084085994148, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.488166809082031, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8606243133544922, "num_tokens": 455932106.0, "step": 11948 }, { "epoch": 1.5200356188780053, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.51253604888916, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8742391467094421, "num_tokens": 455971555.0, "step": 11949 }, { "epoch": 1.5201628291565958, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.533202171325684, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8634766340255737, "num_tokens": 456014177.0, "step": 11950 }, { "epoch": 1.5202900394351864, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.539471626281738, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8737097382545471, "num_tokens": 456048229.0, "step": 11951 }, { "epoch": 1.520417249713777, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.487918853759766, "learning_rate": 1e-06, "loss": 0.3499, "mean_token_accuracy": 0.8872933387756348, "num_tokens": 456085815.0, "step": 11952 }, { "epoch": 1.5205444599923674, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.515768051147461, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8709735870361328, "num_tokens": 456128971.0, "step": 11953 }, { "epoch": 1.520671670270958, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.565951347351074, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8516790866851807, "num_tokens": 456159705.0, "step": 11954 }, { "epoch": 1.5207988805495485, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.504752159118652, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8597891330718994, "num_tokens": 456198251.0, "step": 11955 }, { "epoch": 1.520926090828139, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.516109466552734, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.86103355884552, "num_tokens": 456237612.0, "step": 11956 }, { "epoch": 1.5210533011067295, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.492480278015137, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.86763596534729, "num_tokens": 456273665.0, "step": 11957 }, { "epoch": 1.52118051138532, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.491453170776367, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8696767091751099, "num_tokens": 456305467.0, "step": 11958 }, { "epoch": 1.5213077216639106, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.556678771972656, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8751544952392578, "num_tokens": 456342951.0, "step": 11959 }, { "epoch": 1.521434931942501, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.518485069274902, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8675989508628845, "num_tokens": 456375846.0, "step": 11960 }, { "epoch": 1.5215621422210914, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.542258262634277, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8780204057693481, "num_tokens": 456407993.0, "step": 11961 }, { "epoch": 1.521689352499682, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.497026443481445, "learning_rate": 1e-06, "loss": 0.4699, "mean_token_accuracy": 0.8487998247146606, "num_tokens": 456444278.0, "step": 11962 }, { "epoch": 1.5218165627782725, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.475870132446289, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8772919178009033, "num_tokens": 456483950.0, "step": 11963 }, { "epoch": 1.521943773056863, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.568294525146484, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8688347339630127, "num_tokens": 456527438.0, "step": 11964 }, { "epoch": 1.5220709833354535, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.518449783325195, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8780606985092163, "num_tokens": 456560572.0, "step": 11965 }, { "epoch": 1.5221981936140438, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.522611618041992, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.859811007976532, "num_tokens": 456604781.0, "step": 11966 }, { "epoch": 1.5223254038926344, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.518064498901367, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8819347023963928, "num_tokens": 456639415.0, "step": 11967 }, { "epoch": 1.522452614171225, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.527301788330078, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8814276456832886, "num_tokens": 456671183.0, "step": 11968 }, { "epoch": 1.5225798244498154, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.484604835510254, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8563537001609802, "num_tokens": 456712995.0, "step": 11969 }, { "epoch": 1.522707034728406, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.540020942687988, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8640210628509521, "num_tokens": 456750587.0, "step": 11970 }, { "epoch": 1.5228342450069965, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.535971641540527, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8762706518173218, "num_tokens": 456787503.0, "step": 11971 }, { "epoch": 1.522961455285587, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.528142929077148, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8468135595321655, "num_tokens": 456831645.0, "step": 11972 }, { "epoch": 1.5230886655641775, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.496783256530762, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8751146793365479, "num_tokens": 456869143.0, "step": 11973 }, { "epoch": 1.523215875842768, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.550806999206543, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8670715689659119, "num_tokens": 456906465.0, "step": 11974 }, { "epoch": 1.5233430861213586, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.522790908813477, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8673622608184814, "num_tokens": 456942048.0, "step": 11975 }, { "epoch": 1.5234702963999491, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.540905952453613, "learning_rate": 1e-06, "loss": 0.3519, "mean_token_accuracy": 0.8861976265907288, "num_tokens": 456975710.0, "step": 11976 }, { "epoch": 1.5235975066785397, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.528085708618164, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8576945662498474, "num_tokens": 457022295.0, "step": 11977 }, { "epoch": 1.5237247169571302, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.52845287322998, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8608921766281128, "num_tokens": 457060433.0, "step": 11978 }, { "epoch": 1.5238519272357207, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.529367446899414, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8750213384628296, "num_tokens": 457101760.0, "step": 11979 }, { "epoch": 1.5239791375143112, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.502976417541504, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8843072652816772, "num_tokens": 457142676.0, "step": 11980 }, { "epoch": 1.5241063477929018, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.469829559326172, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8758103251457214, "num_tokens": 457179507.0, "step": 11981 }, { "epoch": 1.5242335580714923, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.451704025268555, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8698009848594666, "num_tokens": 457221563.0, "step": 11982 }, { "epoch": 1.5243607683500828, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.549338340759277, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8660827875137329, "num_tokens": 457258634.0, "step": 11983 }, { "epoch": 1.5244879786286734, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.476309776306152, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8828296661376953, "num_tokens": 457293097.0, "step": 11984 }, { "epoch": 1.5246151889072637, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.521442413330078, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8835240602493286, "num_tokens": 457333525.0, "step": 11985 }, { "epoch": 1.5247423991858542, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.526965141296387, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8543708324432373, "num_tokens": 457379772.0, "step": 11986 }, { "epoch": 1.5248696094644447, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.489423751831055, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8749786019325256, "num_tokens": 457415852.0, "step": 11987 }, { "epoch": 1.5249968197430352, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.504343032836914, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8744077682495117, "num_tokens": 457449771.0, "step": 11988 }, { "epoch": 1.5251240300216258, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.516613960266113, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8648232221603394, "num_tokens": 457489601.0, "step": 11989 }, { "epoch": 1.5252512403002163, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.53141975402832, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8833246231079102, "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": 12.475337028503418, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8655306696891785, "num_tokens": 457568299.0, "step": 11991 }, { "epoch": 1.5255056608573971, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.552467346191406, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.868672788143158, "num_tokens": 457600476.0, "step": 11992 }, { "epoch": 1.5256328711359877, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.501108169555664, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8888254165649414, "num_tokens": 457632024.0, "step": 11993 }, { "epoch": 1.5257600814145782, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.504106521606445, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8752514719963074, "num_tokens": 457670805.0, "step": 11994 }, { "epoch": 1.5258872916931687, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.547256469726562, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8696356415748596, "num_tokens": 457702468.0, "step": 11995 }, { "epoch": 1.5260145019717593, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.488831520080566, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8778342008590698, "num_tokens": 457740443.0, "step": 11996 }, { "epoch": 1.5261417122503498, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.553739547729492, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8715716600418091, "num_tokens": 457775833.0, "step": 11997 }, { "epoch": 1.5262689225289403, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.529274940490723, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8734045028686523, "num_tokens": 457815850.0, "step": 11998 }, { "epoch": 1.5263961328075308, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.499763488769531, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.876316487789154, "num_tokens": 457854946.0, "step": 11999 }, { "epoch": 1.5265233430861214, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.489636421203613, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8743875026702881, "num_tokens": 457890775.0, "step": 12000 }, { "epoch": 1.526650553364712, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.495842933654785, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8744686841964722, "num_tokens": 457928143.0, "step": 12001 }, { "epoch": 1.5267777636433024, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53909969329834, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8750160932540894, "num_tokens": 457965999.0, "step": 12002 }, { "epoch": 1.526904973921893, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.443673133850098, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8692984580993652, "num_tokens": 458006638.0, "step": 12003 }, { "epoch": 1.5270321842004835, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53144359588623, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8698567152023315, "num_tokens": 458047522.0, "step": 12004 }, { "epoch": 1.527159394479074, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.4710054397583, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8701876997947693, "num_tokens": 458091219.0, "step": 12005 }, { "epoch": 1.5272866047576645, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.519600868225098, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8588718175888062, "num_tokens": 458129208.0, "step": 12006 }, { "epoch": 1.527413815036255, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.507370948791504, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8652304410934448, "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": 12.504034996032715, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8636759519577026, "num_tokens": 458202377.0, "step": 12008 }, { "epoch": 1.527668235593436, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.524428367614746, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8694134950637817, "num_tokens": 458239901.0, "step": 12009 }, { "epoch": 1.5277954458720264, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.577743530273438, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.864060640335083, "num_tokens": 458280291.0, "step": 12010 }, { "epoch": 1.527922656150617, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.476385116577148, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8748897314071655, "num_tokens": 458316931.0, "step": 12011 }, { "epoch": 1.5280498664292075, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.45523452758789, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.862818717956543, "num_tokens": 458356477.0, "step": 12012 }, { "epoch": 1.528177076707798, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.565604209899902, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8545271158218384, "num_tokens": 458401340.0, "step": 12013 }, { "epoch": 1.5283042869863885, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.497889518737793, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8763843774795532, "num_tokens": 458436190.0, "step": 12014 }, { "epoch": 1.5284314972649788, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.541216850280762, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8689663410186768, "num_tokens": 458465513.0, "step": 12015 }, { "epoch": 1.5285587075435694, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.486047744750977, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.866112232208252, "num_tokens": 458505465.0, "step": 12016 }, { "epoch": 1.52868591782216, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.513763427734375, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8678252100944519, "num_tokens": 458545087.0, "step": 12017 }, { "epoch": 1.5288131281007504, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.49975872039795, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8767831921577454, "num_tokens": 458583274.0, "step": 12018 }, { "epoch": 1.528940338379341, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.594609260559082, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8633412718772888, "num_tokens": 458626026.0, "step": 12019 }, { "epoch": 1.5290675486579315, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.441810607910156, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8809435963630676, "num_tokens": 458668774.0, "step": 12020 }, { "epoch": 1.529194758936522, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.554709434509277, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8730545043945312, "num_tokens": 458702256.0, "step": 12021 }, { "epoch": 1.5293219692151125, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.52629280090332, "learning_rate": 1e-06, "loss": 0.4939, "mean_token_accuracy": 0.8427039980888367, "num_tokens": 458741429.0, "step": 12022 }, { "epoch": 1.529449179493703, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.490859985351562, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8745706677436829, "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": 12.495747566223145, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8733755350112915, "num_tokens": 458811451.0, "step": 12024 }, { "epoch": 1.5297036000508841, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.455135345458984, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8740721940994263, "num_tokens": 458851761.0, "step": 12025 }, { "epoch": 1.5298308103294747, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.503955841064453, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8634105324745178, "num_tokens": 458888948.0, "step": 12026 }, { "epoch": 1.5299580206080652, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.5081148147583, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8676482439041138, "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": 12.452178001403809, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8645948767662048, "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": 12.4475736618042, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8872774839401245, "num_tokens": 458993634.0, "step": 12029 }, { "epoch": 1.5303396514438368, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.44873046875, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8790396451950073, "num_tokens": 459034314.0, "step": 12030 }, { "epoch": 1.5304668617224273, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.483599662780762, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8750330805778503, "num_tokens": 459072117.0, "step": 12031 }, { "epoch": 1.5305940720010178, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.480504035949707, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8645767569541931, "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": 12.447720527648926, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8818491101264954, "num_tokens": 459148973.0, "step": 12033 }, { "epoch": 1.5308484925581987, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.605355262756348, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8665130138397217, "num_tokens": 459183242.0, "step": 12034 }, { "epoch": 1.5309757028367892, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.503609657287598, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8865441083908081, "num_tokens": 459219606.0, "step": 12035 }, { "epoch": 1.5311029131153797, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.549860000610352, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8873506188392639, "num_tokens": 459253763.0, "step": 12036 }, { "epoch": 1.5312301233939702, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.53211784362793, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.872776985168457, "num_tokens": 459291187.0, "step": 12037 }, { "epoch": 1.5313573336725608, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.52563190460205, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8669041395187378, "num_tokens": 459327423.0, "step": 12038 }, { "epoch": 1.5314845439511513, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.540545463562012, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8652546405792236, "num_tokens": 459366207.0, "step": 12039 }, { "epoch": 1.5316117542297416, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.513538360595703, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8619433641433716, "num_tokens": 459408338.0, "step": 12040 }, { "epoch": 1.5317389645083321, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.495835304260254, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8799796104431152, "num_tokens": 459443420.0, "step": 12041 }, { "epoch": 1.5318661747869227, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.512295722961426, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8548459410667419, "num_tokens": 459475835.0, "step": 12042 }, { "epoch": 1.5319933850655132, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.502829551696777, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8683973550796509, "num_tokens": 459511993.0, "step": 12043 }, { "epoch": 1.5321205953441037, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.521581649780273, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8773361444473267, "num_tokens": 459546221.0, "step": 12044 }, { "epoch": 1.5322478056226942, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.4818696975708, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8739185929298401, "num_tokens": 459587271.0, "step": 12045 }, { "epoch": 1.5323750159012848, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.524961471557617, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8587851524353027, "num_tokens": 459623388.0, "step": 12046 }, { "epoch": 1.5325022261798753, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.507984161376953, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8698257207870483, "num_tokens": 459666440.0, "step": 12047 }, { "epoch": 1.5326294364584658, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.499768257141113, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8567345142364502, "num_tokens": 459708373.0, "step": 12048 }, { "epoch": 1.5327566467370564, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.540877342224121, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.880687952041626, "num_tokens": 459746707.0, "step": 12049 }, { "epoch": 1.532883857015647, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.479750633239746, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8757093548774719, "num_tokens": 459792493.0, "step": 12050 }, { "epoch": 1.5330110672942374, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.614810943603516, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8786650896072388, "num_tokens": 459828897.0, "step": 12051 }, { "epoch": 1.533138277572828, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.509031295776367, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8619742393493652, "num_tokens": 459866274.0, "step": 12052 }, { "epoch": 1.5332654878514185, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.533931732177734, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8855403661727905, "num_tokens": 459902442.0, "step": 12053 }, { "epoch": 1.533392698130009, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.500323295593262, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8777250647544861, "num_tokens": 459943331.0, "step": 12054 }, { "epoch": 1.5335199084085995, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.545401573181152, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8752765655517578, "num_tokens": 459975349.0, "step": 12055 }, { "epoch": 1.53364711868719, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.52920913696289, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8827986717224121, "num_tokens": 460008295.0, "step": 12056 }, { "epoch": 1.5337743289657806, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.492818832397461, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8750292658805847, "num_tokens": 460045892.0, "step": 12057 }, { "epoch": 1.533901539244371, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.542685508728027, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8698712587356567, "num_tokens": 460084255.0, "step": 12058 }, { "epoch": 1.5340287495229614, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.54166316986084, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8561486005783081, "num_tokens": 460123706.0, "step": 12059 }, { "epoch": 1.534155959801552, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.51928424835205, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8563377857208252, "num_tokens": 460162351.0, "step": 12060 }, { "epoch": 1.5342831700801425, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.54399585723877, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8606672883033752, "num_tokens": 460200135.0, "step": 12061 }, { "epoch": 1.534410380358733, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.462385177612305, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8639729022979736, "num_tokens": 460243585.0, "step": 12062 }, { "epoch": 1.5345375906373235, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.550789833068848, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8638565540313721, "num_tokens": 460283695.0, "step": 12063 }, { "epoch": 1.5346648009159138, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.487159729003906, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8763023018836975, "num_tokens": 460317241.0, "step": 12064 }, { "epoch": 1.5347920111945044, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.580926895141602, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8718758225440979, "num_tokens": 460351900.0, "step": 12065 }, { "epoch": 1.534919221473095, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.410773277282715, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.871646523475647, "num_tokens": 460391672.0, "step": 12066 }, { "epoch": 1.5350464317516854, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.542462348937988, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8632365465164185, "num_tokens": 460434120.0, "step": 12067 }, { "epoch": 1.535173642030276, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.481704711914062, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8719834089279175, "num_tokens": 460474214.0, "step": 12068 }, { "epoch": 1.5353008523088665, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.549690246582031, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8761622905731201, "num_tokens": 460507549.0, "step": 12069 }, { "epoch": 1.535428062587457, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.516936302185059, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8629110455513, "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": 12.534043312072754, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8588484525680542, "num_tokens": 460585232.0, "step": 12071 }, { "epoch": 1.535682483144638, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.429438591003418, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8778080940246582, "num_tokens": 460624398.0, "step": 12072 }, { "epoch": 1.5358096934232286, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.563044548034668, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8629032373428345, "num_tokens": 460667193.0, "step": 12073 }, { "epoch": 1.5359369037018191, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.555954933166504, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8469184637069702, "num_tokens": 460710614.0, "step": 12074 }, { "epoch": 1.5360641139804097, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.500844955444336, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8662919998168945, "num_tokens": 460751079.0, "step": 12075 }, { "epoch": 1.5361913242590002, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.537443161010742, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8698039650917053, "num_tokens": 460786768.0, "step": 12076 }, { "epoch": 1.5363185345375907, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.538003921508789, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8601242303848267, "num_tokens": 460824757.0, "step": 12077 }, { "epoch": 1.5364457448161812, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.56279468536377, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8626922369003296, "num_tokens": 460865024.0, "step": 12078 }, { "epoch": 1.5365729550947718, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.55894947052002, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8613055348396301, "num_tokens": 460900895.0, "step": 12079 }, { "epoch": 1.5367001653733623, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.498900413513184, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8746062517166138, "num_tokens": 460936876.0, "step": 12080 }, { "epoch": 1.5368273756519528, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.558713912963867, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8540787100791931, "num_tokens": 460975297.0, "step": 12081 }, { "epoch": 1.5369545859305433, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.561755180358887, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8574559688568115, "num_tokens": 461008603.0, "step": 12082 }, { "epoch": 1.5370817962091337, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.553926467895508, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8607921600341797, "num_tokens": 461047770.0, "step": 12083 }, { "epoch": 1.5372090064877242, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.605192184448242, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8565186858177185, "num_tokens": 461088554.0, "step": 12084 }, { "epoch": 1.5373362167663147, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.561649322509766, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8476508855819702, "num_tokens": 461126070.0, "step": 12085 }, { "epoch": 1.5374634270449052, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.54417896270752, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8669375777244568, "num_tokens": 461166394.0, "step": 12086 }, { "epoch": 1.5375906373234958, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.566791534423828, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8612867593765259, "num_tokens": 461208263.0, "step": 12087 }, { "epoch": 1.5377178476020863, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.527708053588867, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8787931203842163, "num_tokens": 461245054.0, "step": 12088 }, { "epoch": 1.5378450578806766, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.606086730957031, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8579815626144409, "num_tokens": 461281584.0, "step": 12089 }, { "epoch": 1.5379722681592671, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.526788711547852, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8775337338447571, "num_tokens": 461322247.0, "step": 12090 }, { "epoch": 1.5380994784378577, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.584209442138672, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8572087287902832, "num_tokens": 461358371.0, "step": 12091 }, { "epoch": 1.5382266887164482, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.500887870788574, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8748708367347717, "num_tokens": 461394666.0, "step": 12092 }, { "epoch": 1.5383538989950387, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.567314147949219, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8812904357910156, "num_tokens": 461430875.0, "step": 12093 }, { "epoch": 1.5384811092736292, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.524276733398438, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8655139207839966, "num_tokens": 461467412.0, "step": 12094 }, { "epoch": 1.5386083195522198, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.542217254638672, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8471089601516724, "num_tokens": 461501970.0, "step": 12095 }, { "epoch": 1.5387355298308103, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.588629722595215, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8563892841339111, "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": 12.55654239654541, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8708696365356445, "num_tokens": 461583883.0, "step": 12097 }, { "epoch": 1.5389899503879914, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.580487251281738, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8613157272338867, "num_tokens": 461625124.0, "step": 12098 }, { "epoch": 1.5391171606665819, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.574188232421875, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8592013120651245, "num_tokens": 461660273.0, "step": 12099 }, { "epoch": 1.5392443709451724, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.557886123657227, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8632057905197144, "num_tokens": 461698890.0, "step": 12100 }, { "epoch": 1.539371581223763, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.512521743774414, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8681372404098511, "num_tokens": 461736927.0, "step": 12101 }, { "epoch": 1.5394987915023535, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.581789016723633, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.866573691368103, "num_tokens": 461777796.0, "step": 12102 }, { "epoch": 1.539626001780944, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.560989379882812, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8775973320007324, "num_tokens": 461814656.0, "step": 12103 }, { "epoch": 1.5397532120595345, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.60542106628418, "learning_rate": 1e-06, "loss": 0.4752, "mean_token_accuracy": 0.8516870141029358, "num_tokens": 461852186.0, "step": 12104 }, { "epoch": 1.539880422338125, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.49504566192627, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8820387125015259, "num_tokens": 461893724.0, "step": 12105 }, { "epoch": 1.5400076326167156, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.593448638916016, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8558175563812256, "num_tokens": 461935380.0, "step": 12106 }, { "epoch": 1.5401348428953059, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.552510261535645, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8744163513183594, "num_tokens": 461971463.0, "step": 12107 }, { "epoch": 1.5402620531738964, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.55298137664795, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8752403259277344, "num_tokens": 462007514.0, "step": 12108 }, { "epoch": 1.540389263452487, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.532842636108398, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8811358213424683, "num_tokens": 462046022.0, "step": 12109 }, { "epoch": 1.5405164737310775, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.510688781738281, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8733566999435425, "num_tokens": 462089367.0, "step": 12110 }, { "epoch": 1.540643684009668, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.563329696655273, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8745905160903931, "num_tokens": 462128558.0, "step": 12111 }, { "epoch": 1.5407708942882585, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.51506519317627, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8761423826217651, "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": 12.557843208312988, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8694869875907898, "num_tokens": 462199303.0, "step": 12113 }, { "epoch": 1.5410253148454394, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.523873329162598, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8662746548652649, "num_tokens": 462246095.0, "step": 12114 }, { "epoch": 1.54115252512403, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.46187686920166, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8654321432113647, "num_tokens": 462286653.0, "step": 12115 }, { "epoch": 1.5412797354026204, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.530674934387207, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8656007051467896, "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": 12.613277435302734, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8705210089683533, "num_tokens": 462365175.0, "step": 12117 }, { "epoch": 1.5415341559598015, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.522171974182129, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8619468808174133, "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": 12.505271911621094, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.862950325012207, "num_tokens": 462443819.0, "step": 12119 }, { "epoch": 1.5417885765169825, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.538010597229004, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8779268264770508, "num_tokens": 462487319.0, "step": 12120 }, { "epoch": 1.541915786795573, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.568004608154297, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8702183961868286, "num_tokens": 462527026.0, "step": 12121 }, { "epoch": 1.5420429970741636, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.519591331481934, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8595744371414185, "num_tokens": 462557552.0, "step": 12122 }, { "epoch": 1.5421702073527541, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.533393859863281, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8699808120727539, "num_tokens": 462595243.0, "step": 12123 }, { "epoch": 1.5422974176313446, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.49363899230957, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8707876801490784, "num_tokens": 462636077.0, "step": 12124 }, { "epoch": 1.5424246279099352, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.570402145385742, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8811648488044739, "num_tokens": 462668407.0, "step": 12125 }, { "epoch": 1.5425518381885257, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.564603805541992, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8689284324645996, "num_tokens": 462707678.0, "step": 12126 }, { "epoch": 1.5426790484671162, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.595699310302734, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8609045743942261, "num_tokens": 462739722.0, "step": 12127 }, { "epoch": 1.5428062587457068, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.584389686584473, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8762242794036865, "num_tokens": 462770369.0, "step": 12128 }, { "epoch": 1.5429334690242973, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.520966529846191, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8752684593200684, "num_tokens": 462806126.0, "step": 12129 }, { "epoch": 1.5430606793028878, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.604147911071777, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8707801699638367, "num_tokens": 462846585.0, "step": 12130 }, { "epoch": 1.5431878895814783, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.499682426452637, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8653064370155334, "num_tokens": 462889892.0, "step": 12131 }, { "epoch": 1.5433150998600687, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.621557235717773, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8691663146018982, "num_tokens": 462927612.0, "step": 12132 }, { "epoch": 1.5434423101386592, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.583525657653809, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.871812105178833, "num_tokens": 462971661.0, "step": 12133 }, { "epoch": 1.5435695204172497, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.539753913879395, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8743199110031128, "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": 12.622895240783691, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8755478858947754, "num_tokens": 463045395.0, "step": 12135 }, { "epoch": 1.5438239409744308, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.542712211608887, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8589785695075989, "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": 12.575937271118164, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.868433952331543, "num_tokens": 463115658.0, "step": 12137 }, { "epoch": 1.5440783615316116, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.53375244140625, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8810414671897888, "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": 12.623849868774414, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.859587550163269, "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": 12.542326927185059, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8683305978775024, "num_tokens": 463227626.0, "step": 12140 }, { "epoch": 1.5444599923673832, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.573442459106445, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8695408701896667, "num_tokens": 463266451.0, "step": 12141 }, { "epoch": 1.5445872026459737, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.56258487701416, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8575592041015625, "num_tokens": 463305327.0, "step": 12142 }, { "epoch": 1.5447144129245642, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.584081649780273, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8655197620391846, "num_tokens": 463345890.0, "step": 12143 }, { "epoch": 1.5448416232031548, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.554739952087402, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8734076023101807, "num_tokens": 463381583.0, "step": 12144 }, { "epoch": 1.5449688334817453, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.573698997497559, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8770192861557007, "num_tokens": 463414442.0, "step": 12145 }, { "epoch": 1.5450960437603358, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.585773468017578, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8749880194664001, "num_tokens": 463449517.0, "step": 12146 }, { "epoch": 1.5452232540389264, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.515220642089844, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8754552602767944, "num_tokens": 463487952.0, "step": 12147 }, { "epoch": 1.5453504643175169, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.594606399536133, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8694863319396973, "num_tokens": 463520724.0, "step": 12148 }, { "epoch": 1.5454776745961074, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.591261863708496, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8664810657501221, "num_tokens": 463564691.0, "step": 12149 }, { "epoch": 1.545604884874698, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.597930908203125, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.865795373916626, "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": 12.60183334350586, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8564597368240356, "num_tokens": 463630744.0, "step": 12151 }, { "epoch": 1.545859305431879, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.633408546447754, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8653838634490967, "num_tokens": 463663551.0, "step": 12152 }, { "epoch": 1.5459865157104695, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.576488494873047, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8803288340568542, "num_tokens": 463702446.0, "step": 12153 }, { "epoch": 1.54611372598906, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.607369422912598, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8607476949691772, "num_tokens": 463739324.0, "step": 12154 }, { "epoch": 1.5462409362676506, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.640052795410156, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8498324155807495, "num_tokens": 463780765.0, "step": 12155 }, { "epoch": 1.5463681465462409, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.547384262084961, "learning_rate": 1e-06, "loss": 0.4928, "mean_token_accuracy": 0.8477770090103149, "num_tokens": 463817727.0, "step": 12156 }, { "epoch": 1.5464953568248314, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.536262512207031, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8838586211204529, "num_tokens": 463865793.0, "step": 12157 }, { "epoch": 1.546622567103422, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.552149772644043, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8703177571296692, "num_tokens": 463908299.0, "step": 12158 }, { "epoch": 1.5467497773820125, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.620837211608887, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8658623695373535, "num_tokens": 463944513.0, "step": 12159 }, { "epoch": 1.546876987660603, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.598944664001465, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.86183100938797, "num_tokens": 463982862.0, "step": 12160 }, { "epoch": 1.5470041979391935, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.581421852111816, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8726500868797302, "num_tokens": 464023357.0, "step": 12161 }, { "epoch": 1.5471314082177838, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.554697036743164, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8683534860610962, "num_tokens": 464071889.0, "step": 12162 }, { "epoch": 1.5472586184963744, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.58000659942627, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8561217188835144, "num_tokens": 464112334.0, "step": 12163 }, { "epoch": 1.5473858287749649, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.54345417022705, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8641776442527771, "num_tokens": 464156833.0, "step": 12164 }, { "epoch": 1.5475130390535554, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.542423248291016, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8727015256881714, "num_tokens": 464192989.0, "step": 12165 }, { "epoch": 1.547640249332146, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.560416221618652, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8745435476303101, "num_tokens": 464232855.0, "step": 12166 }, { "epoch": 1.5477674596107365, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.576813697814941, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8527639508247375, "num_tokens": 464271860.0, "step": 12167 }, { "epoch": 1.547894669889327, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.560179710388184, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8809686899185181, "num_tokens": 464312925.0, "step": 12168 }, { "epoch": 1.5480218801679175, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.60156536102295, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8642290830612183, "num_tokens": 464349190.0, "step": 12169 }, { "epoch": 1.548149090446508, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.561551094055176, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8636692762374878, "num_tokens": 464392031.0, "step": 12170 }, { "epoch": 1.5482763007250986, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.620221138000488, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8653819561004639, "num_tokens": 464424884.0, "step": 12171 }, { "epoch": 1.5484035110036891, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.571721076965332, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8633279800415039, "num_tokens": 464466047.0, "step": 12172 }, { "epoch": 1.5485307212822796, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.613136291503906, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8673709630966187, "num_tokens": 464497293.0, "step": 12173 }, { "epoch": 1.5486579315608702, "ewc_loss": 0.029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 12.54690933227539, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8552088737487793, "num_tokens": 464541873.0, "step": 12174 }, { "epoch": 1.5487851418394607, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.575338363647461, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8694158792495728, "num_tokens": 464579712.0, "step": 12175 }, { "epoch": 1.5489123521180512, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.51587963104248, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8667043447494507, "num_tokens": 464612277.0, "step": 12176 }, { "epoch": 1.5490395623966418, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.599475860595703, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8651927709579468, "num_tokens": 464647062.0, "step": 12177 }, { "epoch": 1.5491667726752323, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.570661544799805, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8709859848022461, "num_tokens": 464677998.0, "step": 12178 }, { "epoch": 1.5492939829538228, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.547785758972168, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8618218898773193, "num_tokens": 464719205.0, "step": 12179 }, { "epoch": 1.5494211932324133, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.512408256530762, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8729079961776733, "num_tokens": 464760758.0, "step": 12180 }, { "epoch": 1.5495484035110036, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.59256362915039, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8716863989830017, "num_tokens": 464797422.0, "step": 12181 }, { "epoch": 1.5496756137895942, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.534371376037598, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8700289130210876, "num_tokens": 464832889.0, "step": 12182 }, { "epoch": 1.5498028240681847, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.57423210144043, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.861906886100769, "num_tokens": 464867743.0, "step": 12183 }, { "epoch": 1.5499300343467752, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.530922889709473, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8559826016426086, "num_tokens": 464904666.0, "step": 12184 }, { "epoch": 1.5500572446253658, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.512691497802734, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8723113536834717, "num_tokens": 464942886.0, "step": 12185 }, { "epoch": 1.5501844549039563, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.55775260925293, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8676831126213074, "num_tokens": 464977463.0, "step": 12186 }, { "epoch": 1.5503116651825466, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.5557279586792, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8775202035903931, "num_tokens": 465015732.0, "step": 12187 }, { "epoch": 1.5504388754611371, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.52056884765625, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8778582811355591, "num_tokens": 465060789.0, "step": 12188 }, { "epoch": 1.5505660857397277, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.608673095703125, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.865614652633667, "num_tokens": 465097907.0, "step": 12189 }, { "epoch": 1.5506932960183182, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.567632675170898, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8762457370758057, "num_tokens": 465139401.0, "step": 12190 }, { "epoch": 1.5508205062969087, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 12.591463088989258, "learning_rate": 1e-06, "loss": 0.4904, "mean_token_accuracy": 0.8480060696601868, "num_tokens": 465176531.0, "step": 12191 }, { "epoch": 1.5509477165754992, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.60864543914795, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8777321577072144, "num_tokens": 465216074.0, "step": 12192 }, { "epoch": 1.5510749268540898, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.530174255371094, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8815968632698059, "num_tokens": 465257986.0, "step": 12193 }, { "epoch": 1.5512021371326803, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.65103816986084, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.872040331363678, "num_tokens": 465296038.0, "step": 12194 }, { "epoch": 1.5513293474112708, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.602309226989746, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8455280065536499, "num_tokens": 465334215.0, "step": 12195 }, { "epoch": 1.5514565576898613, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.59170150756836, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8677235841751099, "num_tokens": 465373362.0, "step": 12196 }, { "epoch": 1.5515837679684519, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.590508460998535, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8631889224052429, "num_tokens": 465410181.0, "step": 12197 }, { "epoch": 1.5517109782470424, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.641030311584473, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8760601282119751, "num_tokens": 465443056.0, "step": 12198 }, { "epoch": 1.551838188525633, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.54941177368164, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8718885183334351, "num_tokens": 465484509.0, "step": 12199 }, { "epoch": 1.5519653988042235, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.627911567687988, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8671393394470215, "num_tokens": 465520017.0, "step": 12200 }, { "epoch": 1.552092609082814, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.625853538513184, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8851330876350403, "num_tokens": 465557221.0, "step": 12201 }, { "epoch": 1.5522198193614045, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.580055236816406, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8779544234275818, "num_tokens": 465595455.0, "step": 12202 }, { "epoch": 1.552347029639995, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.562895774841309, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8704857230186462, "num_tokens": 465636436.0, "step": 12203 }, { "epoch": 1.5524742399185856, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.610075950622559, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8657053709030151, "num_tokens": 465668956.0, "step": 12204 }, { "epoch": 1.5526014501971759, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.571808815002441, "learning_rate": 1e-06, "loss": 0.4793, "mean_token_accuracy": 0.8505061864852905, "num_tokens": 465708812.0, "step": 12205 }, { "epoch": 1.5527286604757664, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.629260063171387, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8558056354522705, "num_tokens": 465742610.0, "step": 12206 }, { "epoch": 1.552855870754357, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.524818420410156, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8675808310508728, "num_tokens": 465776468.0, "step": 12207 }, { "epoch": 1.5529830810329475, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.711958885192871, "learning_rate": 1e-06, "loss": 0.3542, "mean_token_accuracy": 0.8854852318763733, "num_tokens": 465809300.0, "step": 12208 }, { "epoch": 1.553110291311538, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.539480209350586, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8710520267486572, "num_tokens": 465849847.0, "step": 12209 }, { "epoch": 1.5532375015901285, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.560833930969238, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8799917697906494, "num_tokens": 465884986.0, "step": 12210 }, { "epoch": 1.5533647118687188, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.6278076171875, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8765613436698914, "num_tokens": 465919710.0, "step": 12211 }, { "epoch": 1.5534919221473094, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.594635963439941, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8617926239967346, "num_tokens": 465956041.0, "step": 12212 }, { "epoch": 1.5536191324258999, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.555782318115234, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8803642988204956, "num_tokens": 465994352.0, "step": 12213 }, { "epoch": 1.5537463427044904, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.51603889465332, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8691730499267578, "num_tokens": 466032493.0, "step": 12214 }, { "epoch": 1.553873552983081, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.549030303955078, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.872247040271759, "num_tokens": 466067048.0, "step": 12215 }, { "epoch": 1.5540007632616715, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.560574531555176, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8472449779510498, "num_tokens": 466106895.0, "step": 12216 }, { "epoch": 1.554127973540262, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.554834365844727, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8545499444007874, "num_tokens": 466146740.0, "step": 12217 }, { "epoch": 1.5542551838188525, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.595136642456055, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8748866319656372, "num_tokens": 466187097.0, "step": 12218 }, { "epoch": 1.554382394097443, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.522482872009277, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8593349456787109, "num_tokens": 466223969.0, "step": 12219 }, { "epoch": 1.5545096043760336, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.577816009521484, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8795572519302368, "num_tokens": 466260127.0, "step": 12220 }, { "epoch": 1.554636814654624, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.541911125183105, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8757534027099609, "num_tokens": 466297219.0, "step": 12221 }, { "epoch": 1.5547640249332146, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.611187934875488, "learning_rate": 1e-06, "loss": 0.4762, "mean_token_accuracy": 0.8503929376602173, "num_tokens": 466329474.0, "step": 12222 }, { "epoch": 1.5548912352118052, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.592942237854004, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8667682409286499, "num_tokens": 466372801.0, "step": 12223 }, { "epoch": 1.5550184454903957, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.558218002319336, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8762555122375488, "num_tokens": 466412950.0, "step": 12224 }, { "epoch": 1.5551456557689862, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.565007209777832, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8760195374488831, "num_tokens": 466448830.0, "step": 12225 }, { "epoch": 1.5552728660475768, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.591388702392578, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8586832880973816, "num_tokens": 466489749.0, "step": 12226 }, { "epoch": 1.5554000763261673, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.62873363494873, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.851771891117096, "num_tokens": 466531320.0, "step": 12227 }, { "epoch": 1.5555272866047578, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.598187446594238, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8610455989837646, "num_tokens": 466570001.0, "step": 12228 }, { "epoch": 1.5556544968833483, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.61760139465332, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8714516162872314, "num_tokens": 466607186.0, "step": 12229 }, { "epoch": 1.5557817071619386, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.5614013671875, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8600906729698181, "num_tokens": 466649480.0, "step": 12230 }, { "epoch": 1.5559089174405292, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.581875801086426, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8748666644096375, "num_tokens": 466688861.0, "step": 12231 }, { "epoch": 1.5560361277191197, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.618800163269043, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8496423959732056, "num_tokens": 466723200.0, "step": 12232 }, { "epoch": 1.5561633379977102, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.518671035766602, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.854615330696106, "num_tokens": 466759639.0, "step": 12233 }, { "epoch": 1.5562905482763008, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.607685089111328, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8628380298614502, "num_tokens": 466795748.0, "step": 12234 }, { "epoch": 1.5564177585548913, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.586817741394043, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8789202570915222, "num_tokens": 466832938.0, "step": 12235 }, { "epoch": 1.5565449688334816, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.618572235107422, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8618221282958984, "num_tokens": 466868710.0, "step": 12236 }, { "epoch": 1.5566721791120721, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.634891510009766, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8660416007041931, "num_tokens": 466903558.0, "step": 12237 }, { "epoch": 1.5567993893906626, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.586698532104492, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8590222597122192, "num_tokens": 466946785.0, "step": 12238 }, { "epoch": 1.5569265996692532, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.597841262817383, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8605313897132874, "num_tokens": 466988563.0, "step": 12239 }, { "epoch": 1.5570538099478437, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.60071086883545, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8617895841598511, "num_tokens": 467025447.0, "step": 12240 }, { "epoch": 1.5571810202264342, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.59915542602539, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.876197874546051, "num_tokens": 467060751.0, "step": 12241 }, { "epoch": 1.5573082305050248, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.578546524047852, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8673878908157349, "num_tokens": 467101889.0, "step": 12242 }, { "epoch": 1.5574354407836153, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.583760261535645, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8705590963363647, "num_tokens": 467136782.0, "step": 12243 }, { "epoch": 1.5575626510622058, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.582195281982422, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8772221207618713, "num_tokens": 467174137.0, "step": 12244 }, { "epoch": 1.5576898613407963, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.592305183410645, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8649161458015442, "num_tokens": 467213919.0, "step": 12245 }, { "epoch": 1.5578170716193869, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.591574668884277, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8755996227264404, "num_tokens": 467254860.0, "step": 12246 }, { "epoch": 1.5579442818979774, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.57785415649414, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8796391487121582, "num_tokens": 467296806.0, "step": 12247 }, { "epoch": 1.558071492176568, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.631223678588867, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8754655718803406, "num_tokens": 467332882.0, "step": 12248 }, { "epoch": 1.5581987024551585, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.557633399963379, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8830341696739197, "num_tokens": 467375088.0, "step": 12249 }, { "epoch": 1.558325912733749, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.61097240447998, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.868650496006012, "num_tokens": 467413000.0, "step": 12250 }, { "epoch": 1.5584531230123395, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.589529037475586, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.882789671421051, "num_tokens": 467450947.0, "step": 12251 }, { "epoch": 1.55858033329093, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.663544654846191, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8586006164550781, "num_tokens": 467493770.0, "step": 12252 }, { "epoch": 1.5587075435695206, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.587282180786133, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8655160665512085, "num_tokens": 467531882.0, "step": 12253 }, { "epoch": 1.5588347538481109, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.589088439941406, "learning_rate": 1e-06, "loss": 0.3192, "mean_token_accuracy": 0.8943048715591431, "num_tokens": 467566059.0, "step": 12254 }, { "epoch": 1.5589619641267014, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.562570571899414, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.883090615272522, "num_tokens": 467603734.0, "step": 12255 }, { "epoch": 1.559089174405292, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.649775505065918, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8761968016624451, "num_tokens": 467641224.0, "step": 12256 }, { "epoch": 1.5592163846838825, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.597009658813477, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8672293424606323, "num_tokens": 467677576.0, "step": 12257 }, { "epoch": 1.559343594962473, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.598282814025879, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8752369284629822, "num_tokens": 467717526.0, "step": 12258 }, { "epoch": 1.5594708052410635, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.571450233459473, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.86406409740448, "num_tokens": 467757591.0, "step": 12259 }, { "epoch": 1.5595980155196538, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.606913566589355, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8551716208457947, "num_tokens": 467792794.0, "step": 12260 }, { "epoch": 1.5597252257982444, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.552882194519043, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8638721704483032, "num_tokens": 467829554.0, "step": 12261 }, { "epoch": 1.5598524360768349, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.614546775817871, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8521767854690552, "num_tokens": 467871403.0, "step": 12262 }, { "epoch": 1.5599796463554254, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.571915626525879, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8495451211929321, "num_tokens": 467906498.0, "step": 12263 }, { "epoch": 1.560106856634016, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.601044654846191, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8669261932373047, "num_tokens": 467946426.0, "step": 12264 }, { "epoch": 1.5602340669126065, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.609457969665527, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8709100484848022, "num_tokens": 467979930.0, "step": 12265 }, { "epoch": 1.560361277191197, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.647123336791992, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8624024391174316, "num_tokens": 468016347.0, "step": 12266 }, { "epoch": 1.5604884874697875, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.5836820602417, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8751897811889648, "num_tokens": 468051446.0, "step": 12267 }, { "epoch": 1.560615697748378, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.577069282531738, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8843845129013062, "num_tokens": 468085713.0, "step": 12268 }, { "epoch": 1.5607429080269686, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.602617263793945, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8810486197471619, "num_tokens": 468117863.0, "step": 12269 }, { "epoch": 1.560870118305559, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.563271522521973, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8644003868103027, "num_tokens": 468153041.0, "step": 12270 }, { "epoch": 1.5609973285841496, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.591809272766113, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8672285079956055, "num_tokens": 468193624.0, "step": 12271 }, { "epoch": 1.5611245388627402, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.584066390991211, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8832563757896423, "num_tokens": 468227091.0, "step": 12272 }, { "epoch": 1.5612517491413307, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.6234130859375, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8640763163566589, "num_tokens": 468265079.0, "step": 12273 }, { "epoch": 1.5613789594199212, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.542762756347656, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.865149736404419, "num_tokens": 468307055.0, "step": 12274 }, { "epoch": 1.5615061696985117, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.601847648620605, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8631654381752014, "num_tokens": 468344429.0, "step": 12275 }, { "epoch": 1.5616333799771023, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.578190803527832, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8676820397377014, "num_tokens": 468379448.0, "step": 12276 }, { "epoch": 1.5617605902556928, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.545594215393066, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8714255690574646, "num_tokens": 468423344.0, "step": 12277 }, { "epoch": 1.5618878005342833, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.617112159729004, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8737006783485413, "num_tokens": 468461657.0, "step": 12278 }, { "epoch": 1.5620150108128736, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.516075134277344, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8832554221153259, "num_tokens": 468500428.0, "step": 12279 }, { "epoch": 1.5621422210914642, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.626472473144531, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8576580286026001, "num_tokens": 468536310.0, "step": 12280 }, { "epoch": 1.5622694313700547, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.585587501525879, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8666598796844482, "num_tokens": 468577954.0, "step": 12281 }, { "epoch": 1.5623966416486452, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.644073486328125, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8600480556488037, "num_tokens": 468612131.0, "step": 12282 }, { "epoch": 1.5625238519272358, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.524652481079102, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8641447424888611, "num_tokens": 468654022.0, "step": 12283 }, { "epoch": 1.5626510622058263, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.633909225463867, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8588775396347046, "num_tokens": 468693120.0, "step": 12284 }, { "epoch": 1.5627782724844166, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.570326805114746, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8679257035255432, "num_tokens": 468727695.0, "step": 12285 }, { "epoch": 1.5629054827630071, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.60013484954834, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8652672171592712, "num_tokens": 468766705.0, "step": 12286 }, { "epoch": 1.5630326930415976, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.587390899658203, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8596235513687134, "num_tokens": 468804555.0, "step": 12287 }, { "epoch": 1.5631599033201882, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.653693199157715, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8642081022262573, "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": 12.56191349029541, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8669412136077881, "num_tokens": 468877574.0, "step": 12289 }, { "epoch": 1.5634143238773692, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.632391929626465, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8796354532241821, "num_tokens": 468914573.0, "step": 12290 }, { "epoch": 1.5635415341559598, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.544767379760742, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.878831684589386, "num_tokens": 468950943.0, "step": 12291 }, { "epoch": 1.5636687444345503, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.648259162902832, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8836513757705688, "num_tokens": 468990007.0, "step": 12292 }, { "epoch": 1.5637959547131408, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.655041694641113, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8718423247337341, "num_tokens": 469022461.0, "step": 12293 }, { "epoch": 1.5639231649917313, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.617998123168945, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8736846446990967, "num_tokens": 469056396.0, "step": 12294 }, { "epoch": 1.5640503752703219, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.56214427947998, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8772600293159485, "num_tokens": 469099464.0, "step": 12295 }, { "epoch": 1.5641775855489124, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.556973457336426, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8770309686660767, "num_tokens": 469139830.0, "step": 12296 }, { "epoch": 1.564304795827503, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.615243911743164, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8675476312637329, "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": 12.562721252441406, "learning_rate": 1e-06, "loss": 0.3339, "mean_token_accuracy": 0.8931918144226074, "num_tokens": 469218727.0, "step": 12298 }, { "epoch": 1.564559216384684, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.587358474731445, "learning_rate": 1e-06, "loss": 0.4826, "mean_token_accuracy": 0.8465542793273926, "num_tokens": 469261186.0, "step": 12299 }, { "epoch": 1.5646864266632745, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.665553092956543, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.877013087272644, "num_tokens": 469301191.0, "step": 12300 }, { "epoch": 1.564813636941865, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.542536735534668, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8736622333526611, "num_tokens": 469340132.0, "step": 12301 }, { "epoch": 1.5649408472204556, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.610451698303223, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8865436315536499, "num_tokens": 469376211.0, "step": 12302 }, { "epoch": 1.5650680574990459, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.627850532531738, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8587377071380615, "num_tokens": 469413961.0, "step": 12303 }, { "epoch": 1.5651952677776364, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.688112258911133, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8692421913146973, "num_tokens": 469449507.0, "step": 12304 }, { "epoch": 1.565322478056227, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.586079597473145, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8557754755020142, "num_tokens": 469486324.0, "step": 12305 }, { "epoch": 1.5654496883348175, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.614181518554688, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8642518520355225, "num_tokens": 469523799.0, "step": 12306 }, { "epoch": 1.565576898613408, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.602415084838867, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8552756309509277, "num_tokens": 469558661.0, "step": 12307 }, { "epoch": 1.5657041088919985, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.608574867248535, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8672045469284058, "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": 12.632490158081055, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8685519695281982, "num_tokens": 469637344.0, "step": 12309 }, { "epoch": 1.5659585294491793, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.620162963867188, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8782362937927246, "num_tokens": 469681127.0, "step": 12310 }, { "epoch": 1.5660857397277699, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.544209480285645, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8670856952667236, "num_tokens": 469724238.0, "step": 12311 }, { "epoch": 1.5662129500063604, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.641400337219238, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8619672656059265, "num_tokens": 469765967.0, "step": 12312 }, { "epoch": 1.566340160284951, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.619363784790039, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.869424045085907, "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": 12.674570083618164, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.870165228843689, "num_tokens": 469835448.0, "step": 12314 }, { "epoch": 1.566594580842132, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.588210105895996, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8596118688583374, "num_tokens": 469876647.0, "step": 12315 }, { "epoch": 1.5667217911207225, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.62263011932373, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8832578063011169, "num_tokens": 469918158.0, "step": 12316 }, { "epoch": 1.566849001399313, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.62800121307373, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8544731140136719, "num_tokens": 469953691.0, "step": 12317 }, { "epoch": 1.5669762116779036, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.614890098571777, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.871006429195404, "num_tokens": 469996174.0, "step": 12318 }, { "epoch": 1.567103421956494, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.670125007629395, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.856245219707489, "num_tokens": 470031371.0, "step": 12319 }, { "epoch": 1.5672306322350846, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.571493148803711, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8727288246154785, "num_tokens": 470068783.0, "step": 12320 }, { "epoch": 1.5673578425136752, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.621685981750488, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8515400886535645, "num_tokens": 470103490.0, "step": 12321 }, { "epoch": 1.5674850527922657, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.61754035949707, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8646820783615112, "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": 12.573630332946777, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8722933530807495, "num_tokens": 470171816.0, "step": 12323 }, { "epoch": 1.5677394733494467, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.633807182312012, "learning_rate": 1e-06, "loss": 0.4819, "mean_token_accuracy": 0.8458722829818726, "num_tokens": 470210517.0, "step": 12324 }, { "epoch": 1.5678666836280373, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.605652809143066, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8565215468406677, "num_tokens": 470247085.0, "step": 12325 }, { "epoch": 1.5679938939066278, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.605441093444824, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8687111735343933, "num_tokens": 470288483.0, "step": 12326 }, { "epoch": 1.5681211041852183, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.620267868041992, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8680458068847656, "num_tokens": 470322173.0, "step": 12327 }, { "epoch": 1.5682483144638086, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.55366039276123, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8693983554840088, "num_tokens": 470356617.0, "step": 12328 }, { "epoch": 1.5683755247423992, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.628480911254883, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8797974586486816, "num_tokens": 470393095.0, "step": 12329 }, { "epoch": 1.5685027350209897, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.545806884765625, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8799464106559753, "num_tokens": 470427785.0, "step": 12330 }, { "epoch": 1.5686299452995802, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.62067985534668, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8630156517028809, "num_tokens": 470469971.0, "step": 12331 }, { "epoch": 1.5687571555781707, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.580822944641113, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8606299161911011, "num_tokens": 470503935.0, "step": 12332 }, { "epoch": 1.5688843658567613, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.600212097167969, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8714087009429932, "num_tokens": 470536014.0, "step": 12333 }, { "epoch": 1.5690115761353516, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.585888862609863, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8811719417572021, "num_tokens": 470575632.0, "step": 12334 }, { "epoch": 1.569138786413942, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.616618156433105, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8785708546638489, "num_tokens": 470611295.0, "step": 12335 }, { "epoch": 1.5692659966925326, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.611162185668945, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8672806024551392, "num_tokens": 470646993.0, "step": 12336 }, { "epoch": 1.5693932069711232, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.673102378845215, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8731379508972168, "num_tokens": 470680253.0, "step": 12337 }, { "epoch": 1.5695204172497137, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.601323127746582, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.866119384765625, "num_tokens": 470716169.0, "step": 12338 }, { "epoch": 1.5696476275283042, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.539925575256348, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8813492059707642, "num_tokens": 470754208.0, "step": 12339 }, { "epoch": 1.5697748378068948, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.644693374633789, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8696526288986206, "num_tokens": 470801329.0, "step": 12340 }, { "epoch": 1.5699020480854853, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.634716987609863, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8639746904373169, "num_tokens": 470838408.0, "step": 12341 }, { "epoch": 1.5700292583640758, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.605488777160645, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8691254258155823, "num_tokens": 470874001.0, "step": 12342 }, { "epoch": 1.5701564686426663, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.592745780944824, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.864372968673706, "num_tokens": 470916870.0, "step": 12343 }, { "epoch": 1.5702836789212569, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.72626781463623, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8765603303909302, "num_tokens": 470953862.0, "step": 12344 }, { "epoch": 1.5704108891998474, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.570740699768066, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8786889314651489, "num_tokens": 470991507.0, "step": 12345 }, { "epoch": 1.570538099478438, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.657936096191406, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8723828792572021, "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": 12.646512985229492, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8822387456893921, "num_tokens": 471063829.0, "step": 12347 }, { "epoch": 1.570792520035619, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.599199295043945, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8587520122528076, "num_tokens": 471107259.0, "step": 12348 }, { "epoch": 1.5709197303142095, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.646821975708008, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8717688322067261, "num_tokens": 471141516.0, "step": 12349 }, { "epoch": 1.5710469405928, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.652475357055664, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8699963092803955, "num_tokens": 471178655.0, "step": 12350 }, { "epoch": 1.5711741508713906, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.58765983581543, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8798717260360718, "num_tokens": 471209437.0, "step": 12351 }, { "epoch": 1.5713013611499809, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.620285034179688, "learning_rate": 1e-06, "loss": 0.3557, "mean_token_accuracy": 0.8839921951293945, "num_tokens": 471249559.0, "step": 12352 }, { "epoch": 1.5714285714285714, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.626547813415527, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8750207424163818, "num_tokens": 471287179.0, "step": 12353 }, { "epoch": 1.571555781707162, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.586852073669434, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8724600076675415, "num_tokens": 471328052.0, "step": 12354 }, { "epoch": 1.5716829919857525, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.614075660705566, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8672130107879639, "num_tokens": 471362414.0, "step": 12355 }, { "epoch": 1.571810202264343, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.66426944732666, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8651955127716064, "num_tokens": 471397910.0, "step": 12356 }, { "epoch": 1.5719374125429335, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.58326244354248, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8833150267601013, "num_tokens": 471436951.0, "step": 12357 }, { "epoch": 1.5720646228215238, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.6760892868042, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8723188042640686, "num_tokens": 471470139.0, "step": 12358 }, { "epoch": 1.5721918331001143, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.565018653869629, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8773254156112671, "num_tokens": 471508591.0, "step": 12359 }, { "epoch": 1.5723190433787049, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.66627311706543, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8696776032447815, "num_tokens": 471549165.0, "step": 12360 }, { "epoch": 1.5724462536572954, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.62859058380127, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8645399808883667, "num_tokens": 471589021.0, "step": 12361 }, { "epoch": 1.572573463935886, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.622321128845215, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8631362915039062, "num_tokens": 471624263.0, "step": 12362 }, { "epoch": 1.5727006742144765, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.630529403686523, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8640377521514893, "num_tokens": 471663678.0, "step": 12363 }, { "epoch": 1.572827884493067, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.55853271484375, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8676836490631104, "num_tokens": 471705887.0, "step": 12364 }, { "epoch": 1.5729550947716575, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.605265617370605, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8772577047348022, "num_tokens": 471739796.0, "step": 12365 }, { "epoch": 1.573082305050248, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.626124382019043, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8656439781188965, "num_tokens": 471774723.0, "step": 12366 }, { "epoch": 1.5732095153288386, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.623712539672852, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8621543645858765, "num_tokens": 471814169.0, "step": 12367 }, { "epoch": 1.573336725607429, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.629969596862793, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8705661296844482, "num_tokens": 471852367.0, "step": 12368 }, { "epoch": 1.5734639358860196, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.605483055114746, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8759033679962158, "num_tokens": 471883438.0, "step": 12369 }, { "epoch": 1.5735911461646102, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.593940734863281, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.864690899848938, "num_tokens": 471921051.0, "step": 12370 }, { "epoch": 1.5737183564432007, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.585129737854004, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8731716275215149, "num_tokens": 471959685.0, "step": 12371 }, { "epoch": 1.5738455667217912, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.617462158203125, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.870130717754364, "num_tokens": 471997345.0, "step": 12372 }, { "epoch": 1.5739727770003817, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.601168632507324, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8754061460494995, "num_tokens": 472034455.0, "step": 12373 }, { "epoch": 1.5740999872789723, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.575480461120605, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8686026334762573, "num_tokens": 472073630.0, "step": 12374 }, { "epoch": 1.5742271975575628, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.544310569763184, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8673602342605591, "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": 12.650470733642578, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8759971857070923, "num_tokens": 472148276.0, "step": 12376 }, { "epoch": 1.5744816181147436, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.629264831542969, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8691198825836182, "num_tokens": 472181665.0, "step": 12377 }, { "epoch": 1.5746088283933342, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.622553825378418, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8637392520904541, "num_tokens": 472222759.0, "step": 12378 }, { "epoch": 1.5747360386719247, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.566521644592285, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8636083602905273, "num_tokens": 472260867.0, "step": 12379 }, { "epoch": 1.5748632489505152, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.588044166564941, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8577595949172974, "num_tokens": 472300068.0, "step": 12380 }, { "epoch": 1.5749904592291057, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.616719245910645, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.880219578742981, "num_tokens": 472336459.0, "step": 12381 }, { "epoch": 1.5751176695076963, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.5883150100708, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8742808103561401, "num_tokens": 472376696.0, "step": 12382 }, { "epoch": 1.5752448797862866, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.584464073181152, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8821476101875305, "num_tokens": 472414491.0, "step": 12383 }, { "epoch": 1.575372090064877, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.564844131469727, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8686915636062622, "num_tokens": 472452188.0, "step": 12384 }, { "epoch": 1.5754993003434676, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.61319351196289, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8678275346755981, "num_tokens": 472489690.0, "step": 12385 }, { "epoch": 1.5756265106220582, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.655169486999512, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8682907819747925, "num_tokens": 472529912.0, "step": 12386 }, { "epoch": 1.5757537209006487, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.604732513427734, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.867551326751709, "num_tokens": 472562534.0, "step": 12387 }, { "epoch": 1.5758809311792392, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.660282135009766, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8672740459442139, "num_tokens": 472598955.0, "step": 12388 }, { "epoch": 1.5760081414578297, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.604256629943848, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8729048371315002, "num_tokens": 472638064.0, "step": 12389 }, { "epoch": 1.5761353517364203, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.664447784423828, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8729773759841919, "num_tokens": 472676823.0, "step": 12390 }, { "epoch": 1.5762625620150108, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.637954711914062, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8693163394927979, "num_tokens": 472718528.0, "step": 12391 }, { "epoch": 1.5763897722936013, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.620464324951172, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8749132752418518, "num_tokens": 472755634.0, "step": 12392 }, { "epoch": 1.5765169825721919, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.58106803894043, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8807134628295898, "num_tokens": 472798060.0, "step": 12393 }, { "epoch": 1.5766441928507824, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.622981071472168, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8740705847740173, "num_tokens": 472837628.0, "step": 12394 }, { "epoch": 1.576771403129373, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.60318374633789, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8697055578231812, "num_tokens": 472877446.0, "step": 12395 }, { "epoch": 1.5768986134079634, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.629667282104492, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8630785942077637, "num_tokens": 472918437.0, "step": 12396 }, { "epoch": 1.577025823686554, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.640170097351074, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.878173291683197, "num_tokens": 472956242.0, "step": 12397 }, { "epoch": 1.5771530339651445, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.625531196594238, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8788544535636902, "num_tokens": 472985620.0, "step": 12398 }, { "epoch": 1.577280244243735, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.677204132080078, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.873971164226532, "num_tokens": 473019517.0, "step": 12399 }, { "epoch": 1.5774074545223256, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.590493202209473, "learning_rate": 1e-06, "loss": 0.3412, "mean_token_accuracy": 0.8897122144699097, "num_tokens": 473054624.0, "step": 12400 }, { "epoch": 1.5775346648009159, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.603523254394531, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8604867458343506, "num_tokens": 473094140.0, "step": 12401 }, { "epoch": 1.5776618750795064, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.606758117675781, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.85148686170578, "num_tokens": 473136886.0, "step": 12402 }, { "epoch": 1.577789085358097, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.649556159973145, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8585041761398315, "num_tokens": 473173267.0, "step": 12403 }, { "epoch": 1.5779162956366874, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.668045997619629, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8500734567642212, "num_tokens": 473216477.0, "step": 12404 }, { "epoch": 1.578043505915278, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.589038848876953, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8620532751083374, "num_tokens": 473256986.0, "step": 12405 }, { "epoch": 1.5781707161938685, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.61162281036377, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8774170875549316, "num_tokens": 473297967.0, "step": 12406 }, { "epoch": 1.5782979264724588, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.585485458374023, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8619762659072876, "num_tokens": 473333580.0, "step": 12407 }, { "epoch": 1.5784251367510493, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.647956848144531, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8744792938232422, "num_tokens": 473370553.0, "step": 12408 }, { "epoch": 1.5785523470296399, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.593724250793457, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8522579669952393, "num_tokens": 473410705.0, "step": 12409 }, { "epoch": 1.5786795573082304, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.62635612487793, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8743956089019775, "num_tokens": 473443315.0, "step": 12410 }, { "epoch": 1.578806767586821, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.684211730957031, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8612755537033081, "num_tokens": 473488767.0, "step": 12411 }, { "epoch": 1.5789339778654115, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.623383522033691, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8844844102859497, "num_tokens": 473525360.0, "step": 12412 }, { "epoch": 1.579061188144002, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.657326698303223, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8616043329238892, "num_tokens": 473560625.0, "step": 12413 }, { "epoch": 1.5791883984225925, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.594172477722168, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.8875041007995605, "num_tokens": 473595484.0, "step": 12414 }, { "epoch": 1.579315608701183, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.612343788146973, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8666486740112305, "num_tokens": 473627269.0, "step": 12415 }, { "epoch": 1.5794428189797736, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.546331405639648, "learning_rate": 1e-06, "loss": 0.4519, "mean_token_accuracy": 0.8542926907539368, "num_tokens": 473669159.0, "step": 12416 }, { "epoch": 1.579570029258364, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.635497093200684, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8789693117141724, "num_tokens": 473707739.0, "step": 12417 }, { "epoch": 1.5796972395369546, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.553105354309082, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8707512617111206, "num_tokens": 473746908.0, "step": 12418 }, { "epoch": 1.5798244498155452, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.681075096130371, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8681689500808716, "num_tokens": 473781480.0, "step": 12419 }, { "epoch": 1.5799516600941357, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.586358070373535, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8685328960418701, "num_tokens": 473814676.0, "step": 12420 }, { "epoch": 1.5800788703727262, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.618844032287598, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8654273748397827, "num_tokens": 473855487.0, "step": 12421 }, { "epoch": 1.5802060806513167, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.657773971557617, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8825981020927429, "num_tokens": 473885648.0, "step": 12422 }, { "epoch": 1.5803332909299073, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.660308837890625, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8653386235237122, "num_tokens": 473923862.0, "step": 12423 }, { "epoch": 1.5804605012084978, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.632473945617676, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8736803531646729, "num_tokens": 473957725.0, "step": 12424 }, { "epoch": 1.5805877114870883, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.602087020874023, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8704265356063843, "num_tokens": 473994939.0, "step": 12425 }, { "epoch": 1.5807149217656786, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.651129722595215, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8655552268028259, "num_tokens": 474035996.0, "step": 12426 }, { "epoch": 1.5808421320442692, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.67665958404541, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8787768483161926, "num_tokens": 474071693.0, "step": 12427 }, { "epoch": 1.5809693423228597, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.612860679626465, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8762018084526062, "num_tokens": 474114415.0, "step": 12428 }, { "epoch": 1.5810965526014502, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.672297477722168, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8638817071914673, "num_tokens": 474152060.0, "step": 12429 }, { "epoch": 1.5812237628800407, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.665374755859375, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8741198778152466, "num_tokens": 474190605.0, "step": 12430 }, { "epoch": 1.5813509731586313, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.599608421325684, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8644431233406067, "num_tokens": 474229447.0, "step": 12431 }, { "epoch": 1.5814781834372216, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.642731666564941, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8767110705375671, "num_tokens": 474268088.0, "step": 12432 }, { "epoch": 1.581605393715812, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.650634765625, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8657544255256653, "num_tokens": 474304856.0, "step": 12433 }, { "epoch": 1.5817326039944026, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.628724098205566, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8725148439407349, "num_tokens": 474349248.0, "step": 12434 }, { "epoch": 1.5818598142729932, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.649721145629883, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8803666830062866, "num_tokens": 474384432.0, "step": 12435 }, { "epoch": 1.5819870245515837, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.649571418762207, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8673256635665894, "num_tokens": 474415881.0, "step": 12436 }, { "epoch": 1.5821142348301742, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.64918327331543, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8675916194915771, "num_tokens": 474452755.0, "step": 12437 }, { "epoch": 1.5822414451087647, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.636775970458984, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.864981472492218, "num_tokens": 474493937.0, "step": 12438 }, { "epoch": 1.5823686553873553, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.644338607788086, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8761099576950073, "num_tokens": 474534072.0, "step": 12439 }, { "epoch": 1.5824958656659458, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.650914192199707, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8619232773780823, "num_tokens": 474569343.0, "step": 12440 }, { "epoch": 1.5826230759445363, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.660136222839355, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8595404624938965, "num_tokens": 474606610.0, "step": 12441 }, { "epoch": 1.5827502862231269, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.604619979858398, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8608696460723877, "num_tokens": 474645071.0, "step": 12442 }, { "epoch": 1.5828774965017174, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.653932571411133, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8563476800918579, "num_tokens": 474683985.0, "step": 12443 }, { "epoch": 1.583004706780308, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.626620292663574, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8751295804977417, "num_tokens": 474726102.0, "step": 12444 }, { "epoch": 1.5831319170588984, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.614113807678223, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8721952438354492, "num_tokens": 474766430.0, "step": 12445 }, { "epoch": 1.583259127337489, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.642861366271973, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8689348697662354, "num_tokens": 474800648.0, "step": 12446 }, { "epoch": 1.5833863376160795, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.607770919799805, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.873003363609314, "num_tokens": 474845540.0, "step": 12447 }, { "epoch": 1.58351354789467, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.674677848815918, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8671618103981018, "num_tokens": 474884296.0, "step": 12448 }, { "epoch": 1.5836407581732606, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.642269134521484, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8735625743865967, "num_tokens": 474929344.0, "step": 12449 }, { "epoch": 1.5837679684518509, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.663652420043945, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8744709491729736, "num_tokens": 474968035.0, "step": 12450 }, { "epoch": 1.5838951787304414, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.692608833312988, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.864029049873352, "num_tokens": 475006189.0, "step": 12451 }, { "epoch": 1.584022389009032, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.633549690246582, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8642427921295166, "num_tokens": 475039768.0, "step": 12452 }, { "epoch": 1.5841495992876224, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.660480499267578, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8750623464584351, "num_tokens": 475079938.0, "step": 12453 }, { "epoch": 1.584276809566213, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.621204376220703, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8574637174606323, "num_tokens": 475122405.0, "step": 12454 }, { "epoch": 1.5844040198448035, "ewc_loss": 0.029541015625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 12.618701934814453, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.862479031085968, "num_tokens": 475158237.0, "step": 12455 }, { "epoch": 1.5845312301233938, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.638382911682129, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8705484867095947, "num_tokens": 475194966.0, "step": 12456 }, { "epoch": 1.5846584404019843, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.6199951171875, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8690327405929565, "num_tokens": 475234483.0, "step": 12457 }, { "epoch": 1.5847856506805749, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.661579132080078, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8641594648361206, "num_tokens": 475275172.0, "step": 12458 }, { "epoch": 1.5849128609591654, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.645166397094727, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8744474053382874, "num_tokens": 475313550.0, "step": 12459 }, { "epoch": 1.585040071237756, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.648037910461426, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8768267631530762, "num_tokens": 475357019.0, "step": 12460 }, { "epoch": 1.5851672815163464, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.640506744384766, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8739221692085266, "num_tokens": 475394788.0, "step": 12461 }, { "epoch": 1.585294491794937, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.653350830078125, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8681869506835938, "num_tokens": 475432354.0, "step": 12462 }, { "epoch": 1.5854217020735275, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.649725914001465, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8736799359321594, "num_tokens": 475467284.0, "step": 12463 }, { "epoch": 1.585548912352118, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.654614448547363, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8625590801239014, "num_tokens": 475505946.0, "step": 12464 }, { "epoch": 1.5856761226307086, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.68785285949707, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8557968139648438, "num_tokens": 475542799.0, "step": 12465 }, { "epoch": 1.585803332909299, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.618264198303223, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8698871731758118, "num_tokens": 475582064.0, "step": 12466 }, { "epoch": 1.5859305431878896, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.725907325744629, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8636853098869324, "num_tokens": 475615521.0, "step": 12467 }, { "epoch": 1.5860577534664801, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.648919105529785, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8776800632476807, "num_tokens": 475650500.0, "step": 12468 }, { "epoch": 1.5861849637450707, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.685708045959473, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8709648847579956, "num_tokens": 475695824.0, "step": 12469 }, { "epoch": 1.5863121740236612, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.605592727661133, "learning_rate": 1e-06, "loss": 0.485, "mean_token_accuracy": 0.8469551801681519, "num_tokens": 475734015.0, "step": 12470 }, { "epoch": 1.5864393843022517, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.668028831481934, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8671368956565857, "num_tokens": 475768961.0, "step": 12471 }, { "epoch": 1.5865665945808423, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.643293380737305, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8830386400222778, "num_tokens": 475806551.0, "step": 12472 }, { "epoch": 1.5866938048594328, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.689570426940918, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8625362515449524, "num_tokens": 475844066.0, "step": 12473 }, { "epoch": 1.5868210151380233, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.682718276977539, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8743073344230652, "num_tokens": 475874554.0, "step": 12474 }, { "epoch": 1.5869482254166136, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.633849143981934, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8757909536361694, "num_tokens": 475915840.0, "step": 12475 }, { "epoch": 1.5870754356952042, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 12.69815731048584, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8679814338684082, "num_tokens": 475951516.0, "step": 12476 }, { "epoch": 1.5872026459737947, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.673652648925781, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8716520071029663, "num_tokens": 475984224.0, "step": 12477 }, { "epoch": 1.5873298562523852, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.668886184692383, "learning_rate": 1e-06, "loss": 0.4713, "mean_token_accuracy": 0.8473656177520752, "num_tokens": 476015758.0, "step": 12478 }, { "epoch": 1.5874570665309757, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.657563209533691, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8655322790145874, "num_tokens": 476051835.0, "step": 12479 }, { "epoch": 1.5875842768095663, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.594245910644531, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8655308485031128, "num_tokens": 476090667.0, "step": 12480 }, { "epoch": 1.5877114870881566, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.628579139709473, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8667178153991699, "num_tokens": 476130448.0, "step": 12481 }, { "epoch": 1.587838697366747, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.614092826843262, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8598918914794922, "num_tokens": 476174495.0, "step": 12482 }, { "epoch": 1.5879659076453376, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.574031829833984, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8779210448265076, "num_tokens": 476213412.0, "step": 12483 }, { "epoch": 1.5880931179239282, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.57054328918457, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8827229738235474, "num_tokens": 476250441.0, "step": 12484 }, { "epoch": 1.5882203282025187, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.61467170715332, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8826038241386414, "num_tokens": 476284466.0, "step": 12485 }, { "epoch": 1.5883475384811092, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.561959266662598, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8684629201889038, "num_tokens": 476327760.0, "step": 12486 }, { "epoch": 1.5884747487596997, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.583148002624512, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8647575378417969, "num_tokens": 476369259.0, "step": 12487 }, { "epoch": 1.5886019590382903, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.63469409942627, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8703352212905884, "num_tokens": 476403915.0, "step": 12488 }, { "epoch": 1.5887291693168808, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.586063385009766, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8671654462814331, "num_tokens": 476440837.0, "step": 12489 }, { "epoch": 1.5888563795954713, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.646035194396973, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8674755096435547, "num_tokens": 476478478.0, "step": 12490 }, { "epoch": 1.5889835898740619, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.611181259155273, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8776007890701294, "num_tokens": 476514675.0, "step": 12491 }, { "epoch": 1.5891108001526524, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.637751579284668, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8668328523635864, "num_tokens": 476559015.0, "step": 12492 }, { "epoch": 1.589238010431243, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.619187355041504, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8543316721916199, "num_tokens": 476601848.0, "step": 12493 }, { "epoch": 1.5893652207098334, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.647664070129395, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8777377605438232, "num_tokens": 476636794.0, "step": 12494 }, { "epoch": 1.589492430988424, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.625506401062012, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8636928200721741, "num_tokens": 476671593.0, "step": 12495 }, { "epoch": 1.5896196412670145, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.640742301940918, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8790000677108765, "num_tokens": 476706776.0, "step": 12496 }, { "epoch": 1.589746851545605, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.629622459411621, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8685291409492493, "num_tokens": 476746700.0, "step": 12497 }, { "epoch": 1.5898740618241956, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.593912124633789, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8738982081413269, "num_tokens": 476785837.0, "step": 12498 }, { "epoch": 1.5900012721027859, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.647518157958984, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8740144371986389, "num_tokens": 476818894.0, "step": 12499 }, { "epoch": 1.5901284823813764, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.61442756652832, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8522584438323975, "num_tokens": 476861773.0, "step": 12500 }, { "epoch": 1.590255692659967, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.632942199707031, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8744473457336426, "num_tokens": 476899522.0, "step": 12501 }, { "epoch": 1.5903829029385574, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.684978485107422, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8632338047027588, "num_tokens": 476939923.0, "step": 12502 }, { "epoch": 1.590510113217148, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.658614158630371, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8685125112533569, "num_tokens": 476979867.0, "step": 12503 }, { "epoch": 1.5906373234957385, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.615495681762695, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8773630857467651, "num_tokens": 477016068.0, "step": 12504 }, { "epoch": 1.5907645337743288, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.677117347717285, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8658464550971985, "num_tokens": 477054526.0, "step": 12505 }, { "epoch": 1.5908917440529193, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.596858978271484, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8645110130310059, "num_tokens": 477098029.0, "step": 12506 }, { "epoch": 1.5910189543315099, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.64151668548584, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8758533000946045, "num_tokens": 477137070.0, "step": 12507 }, { "epoch": 1.5911461646101004, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.633623123168945, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8671454191207886, "num_tokens": 477175640.0, "step": 12508 }, { "epoch": 1.591273374888691, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.671302795410156, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8713234066963196, "num_tokens": 477210935.0, "step": 12509 }, { "epoch": 1.5914005851672814, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.654227256774902, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8758842945098877, "num_tokens": 477251567.0, "step": 12510 }, { "epoch": 1.591527795445872, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.667070388793945, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8713994026184082, "num_tokens": 477287964.0, "step": 12511 }, { "epoch": 1.5916550057244625, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.694674491882324, "learning_rate": 1e-06, "loss": 0.483, "mean_token_accuracy": 0.8439797163009644, "num_tokens": 477328660.0, "step": 12512 }, { "epoch": 1.591782216003053, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.627150535583496, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8740224242210388, "num_tokens": 477364111.0, "step": 12513 }, { "epoch": 1.5919094262816436, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.673184394836426, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8754821419715881, "num_tokens": 477402691.0, "step": 12514 }, { "epoch": 1.592036636560234, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.71300983428955, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8825318217277527, "num_tokens": 477437526.0, "step": 12515 }, { "epoch": 1.5921638468388246, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.607124328613281, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8674415349960327, "num_tokens": 477479175.0, "step": 12516 }, { "epoch": 1.5922910571174151, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.681183815002441, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8712465763092041, "num_tokens": 477516896.0, "step": 12517 }, { "epoch": 1.5924182673960057, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.659197807312012, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8597687482833862, "num_tokens": 477555543.0, "step": 12518 }, { "epoch": 1.5925454776745962, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.620298385620117, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8808518648147583, "num_tokens": 477592251.0, "step": 12519 }, { "epoch": 1.5926726879531867, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.69406509399414, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8562061786651611, "num_tokens": 477630039.0, "step": 12520 }, { "epoch": 1.5927998982317773, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.609979629516602, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.867572546005249, "num_tokens": 477661028.0, "step": 12521 }, { "epoch": 1.5929271085103678, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.65357780456543, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.882800817489624, "num_tokens": 477699612.0, "step": 12522 }, { "epoch": 1.5930543187889583, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.656006813049316, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8650962114334106, "num_tokens": 477737300.0, "step": 12523 }, { "epoch": 1.5931815290675486, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.622736930847168, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8574569225311279, "num_tokens": 477776031.0, "step": 12524 }, { "epoch": 1.5933087393461391, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.683792114257812, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8700721263885498, "num_tokens": 477813640.0, "step": 12525 }, { "epoch": 1.5934359496247297, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.664314270019531, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8649214506149292, "num_tokens": 477848652.0, "step": 12526 }, { "epoch": 1.5935631599033202, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.668784141540527, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8651654720306396, "num_tokens": 477883710.0, "step": 12527 }, { "epoch": 1.5936903701819107, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.655921936035156, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8628295063972473, "num_tokens": 477920316.0, "step": 12528 }, { "epoch": 1.5938175804605013, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.665322303771973, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8782113790512085, "num_tokens": 477955523.0, "step": 12529 }, { "epoch": 1.5939447907390916, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.650681495666504, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8612024188041687, "num_tokens": 477995768.0, "step": 12530 }, { "epoch": 1.594072001017682, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.613447189331055, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8688153028488159, "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": 12.60637378692627, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.857074499130249, "num_tokens": 478077388.0, "step": 12532 }, { "epoch": 1.5943264215748632, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.653404235839844, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8833965063095093, "num_tokens": 478109494.0, "step": 12533 }, { "epoch": 1.5944536318534537, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.644791603088379, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8643043041229248, "num_tokens": 478142420.0, "step": 12534 }, { "epoch": 1.5945808421320442, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.593667984008789, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8737056851387024, "num_tokens": 478183848.0, "step": 12535 }, { "epoch": 1.5947080524106347, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.645307540893555, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8789094686508179, "num_tokens": 478225242.0, "step": 12536 }, { "epoch": 1.5948352626892253, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.607258796691895, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8755427598953247, "num_tokens": 478263167.0, "step": 12537 }, { "epoch": 1.5949624729678158, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.771377563476562, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8699238300323486, "num_tokens": 478297752.0, "step": 12538 }, { "epoch": 1.5950896832464063, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.580341339111328, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8696410059928894, "num_tokens": 478332955.0, "step": 12539 }, { "epoch": 1.5952168935249968, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.744004249572754, "learning_rate": 1e-06, "loss": 0.471, "mean_token_accuracy": 0.8487962484359741, "num_tokens": 478367298.0, "step": 12540 }, { "epoch": 1.5953441038035874, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.64040756225586, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8730305433273315, "num_tokens": 478411183.0, "step": 12541 }, { "epoch": 1.595471314082178, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.685964584350586, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8672051429748535, "num_tokens": 478451184.0, "step": 12542 }, { "epoch": 1.5955985243607684, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.681628227233887, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8651493191719055, "num_tokens": 478490913.0, "step": 12543 }, { "epoch": 1.595725734639359, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.685321807861328, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8699454665184021, "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": 12.668102264404297, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8663274049758911, "num_tokens": 478566665.0, "step": 12545 }, { "epoch": 1.59598015519654, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.679210662841797, "learning_rate": 1e-06, "loss": 0.3267, "mean_token_accuracy": 0.8951008319854736, "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": 12.622413635253906, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.871526837348938, "num_tokens": 478641610.0, "step": 12547 }, { "epoch": 1.5962345757537209, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.640957832336426, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8771335482597351, "num_tokens": 478678183.0, "step": 12548 }, { "epoch": 1.5963617860323114, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.646181106567383, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.870141863822937, "num_tokens": 478710881.0, "step": 12549 }, { "epoch": 1.596488996310902, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.596418380737305, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8835338354110718, "num_tokens": 478748550.0, "step": 12550 }, { "epoch": 1.5966162065894924, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.58232307434082, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8578839302062988, "num_tokens": 478790417.0, "step": 12551 }, { "epoch": 1.596743416868083, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.676875114440918, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8754099607467651, "num_tokens": 478828856.0, "step": 12552 }, { "epoch": 1.5968706271466735, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.630743026733398, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8665392994880676, "num_tokens": 478870240.0, "step": 12553 }, { "epoch": 1.5969978374252638, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.611368179321289, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8615825772285461, "num_tokens": 478911920.0, "step": 12554 }, { "epoch": 1.5971250477038543, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.629500389099121, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8629668354988098, "num_tokens": 478949083.0, "step": 12555 }, { "epoch": 1.5972522579824449, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.664294242858887, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8676616549491882, "num_tokens": 478984087.0, "step": 12556 }, { "epoch": 1.5973794682610354, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.701737403869629, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8508055210113525, "num_tokens": 479019364.0, "step": 12557 }, { "epoch": 1.597506678539626, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.651562690734863, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8730721473693848, "num_tokens": 479059002.0, "step": 12558 }, { "epoch": 1.5976338888182164, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.680814743041992, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8683955669403076, "num_tokens": 479092676.0, "step": 12559 }, { "epoch": 1.597761099096807, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.729340553283691, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8724969029426575, "num_tokens": 479125951.0, "step": 12560 }, { "epoch": 1.5978883093753975, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.616886138916016, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8837271928787231, "num_tokens": 479166230.0, "step": 12561 }, { "epoch": 1.598015519653988, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.703185081481934, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8659452199935913, "num_tokens": 479205813.0, "step": 12562 }, { "epoch": 1.5981427299325786, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.667814254760742, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8654956221580505, "num_tokens": 479244960.0, "step": 12563 }, { "epoch": 1.598269940211169, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.726920127868652, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.87549889087677, "num_tokens": 479289617.0, "step": 12564 }, { "epoch": 1.5983971504897596, "ewc_loss": 0.02978515625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.648374557495117, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8697646856307983, "num_tokens": 479326377.0, "step": 12565 }, { "epoch": 1.5985243607683501, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.664909362792969, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8687339425086975, "num_tokens": 479360563.0, "step": 12566 }, { "epoch": 1.5986515710469407, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.660381317138672, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8805639743804932, "num_tokens": 479403042.0, "step": 12567 }, { "epoch": 1.5987787813255312, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.658990859985352, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8752490878105164, "num_tokens": 479437099.0, "step": 12568 }, { "epoch": 1.5989059916041217, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.678215980529785, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8684285879135132, "num_tokens": 479479326.0, "step": 12569 }, { "epoch": 1.5990332018827123, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.681148529052734, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8525458574295044, "num_tokens": 479515320.0, "step": 12570 }, { "epoch": 1.5991604121613028, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.648392677307129, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8577957153320312, "num_tokens": 479559960.0, "step": 12571 }, { "epoch": 1.5992876224398933, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.67558765411377, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8699049949645996, "num_tokens": 479599115.0, "step": 12572 }, { "epoch": 1.5994148327184836, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.665484428405762, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8773115873336792, "num_tokens": 479636116.0, "step": 12573 }, { "epoch": 1.5995420429970741, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.688613891601562, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8605985045433044, "num_tokens": 479676880.0, "step": 12574 }, { "epoch": 1.5996692532756647, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.66749382019043, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.871878981590271, "num_tokens": 479717424.0, "step": 12575 }, { "epoch": 1.5997964635542552, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.681265830993652, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8675904273986816, "num_tokens": 479756240.0, "step": 12576 }, { "epoch": 1.5999236738328457, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.701108932495117, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8545392155647278, "num_tokens": 479791863.0, "step": 12577 }, { "epoch": 1.6000508841114363, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.617905616760254, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8795143365859985, "num_tokens": 479827252.0, "step": 12578 }, { "epoch": 1.6001780943900266, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.648802757263184, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8658890724182129, "num_tokens": 479863100.0, "step": 12579 }, { "epoch": 1.600305304668617, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.662383079528809, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.859193742275238, "num_tokens": 479899684.0, "step": 12580 }, { "epoch": 1.6004325149472076, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.698556900024414, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8703046441078186, "num_tokens": 479937656.0, "step": 12581 }, { "epoch": 1.6005597252257981, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.653572082519531, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8654949069023132, "num_tokens": 479973357.0, "step": 12582 }, { "epoch": 1.6006869355043887, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.645082473754883, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8700040578842163, "num_tokens": 480017862.0, "step": 12583 }, { "epoch": 1.6008141457829792, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.662088394165039, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.875365138053894, "num_tokens": 480057039.0, "step": 12584 }, { "epoch": 1.6009413560615697, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.654220581054688, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8720225095748901, "num_tokens": 480093396.0, "step": 12585 }, { "epoch": 1.6010685663401603, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.66630744934082, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8654537200927734, "num_tokens": 480125200.0, "step": 12586 }, { "epoch": 1.6011957766187508, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.709802627563477, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8663721084594727, "num_tokens": 480158085.0, "step": 12587 }, { "epoch": 1.6013229868973413, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.589372634887695, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8845689296722412, "num_tokens": 480192979.0, "step": 12588 }, { "epoch": 1.6014501971759318, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.684956550598145, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.879642128944397, "num_tokens": 480227984.0, "step": 12589 }, { "epoch": 1.6015774074545224, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.59548568725586, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8867474794387817, "num_tokens": 480264280.0, "step": 12590 }, { "epoch": 1.601704617733113, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.659956932067871, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8556397557258606, "num_tokens": 480300670.0, "step": 12591 }, { "epoch": 1.6018318280117034, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.673799514770508, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.867554247379303, "num_tokens": 480334460.0, "step": 12592 }, { "epoch": 1.601959038290294, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.651301383972168, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8789109587669373, "num_tokens": 480376066.0, "step": 12593 }, { "epoch": 1.6020862485688845, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.646549224853516, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8754546642303467, "num_tokens": 480417575.0, "step": 12594 }, { "epoch": 1.602213458847475, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.680773735046387, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8594886064529419, "num_tokens": 480454106.0, "step": 12595 }, { "epoch": 1.6023406691260655, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.587773323059082, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.857754111289978, "num_tokens": 480493031.0, "step": 12596 }, { "epoch": 1.6024678794046558, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.705743789672852, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8709516525268555, "num_tokens": 480528326.0, "step": 12597 }, { "epoch": 1.6025950896832464, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.611964225769043, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8777273893356323, "num_tokens": 480565399.0, "step": 12598 }, { "epoch": 1.602722299961837, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.6294527053833, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8850368857383728, "num_tokens": 480600028.0, "step": 12599 }, { "epoch": 1.6028495102404274, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.672479629516602, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8689773678779602, "num_tokens": 480635759.0, "step": 12600 }, { "epoch": 1.602976720519018, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.610311508178711, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8751671314239502, "num_tokens": 480669528.0, "step": 12601 }, { "epoch": 1.6031039307976085, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.670541763305664, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8683048486709595, "num_tokens": 480709425.0, "step": 12602 }, { "epoch": 1.6032311410761988, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.670453071594238, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.856171190738678, "num_tokens": 480745010.0, "step": 12603 }, { "epoch": 1.6033583513547893, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.686599731445312, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8764410018920898, "num_tokens": 480780320.0, "step": 12604 }, { "epoch": 1.6034855616333799, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.684962272644043, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8766360282897949, "num_tokens": 480819335.0, "step": 12605 }, { "epoch": 1.6036127719119704, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.75045394897461, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8825578689575195, "num_tokens": 480852702.0, "step": 12606 }, { "epoch": 1.603739982190561, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.689517974853516, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.864427387714386, "num_tokens": 480890258.0, "step": 12607 }, { "epoch": 1.6038671924691514, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.63852596282959, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.869232177734375, "num_tokens": 480929555.0, "step": 12608 }, { "epoch": 1.603994402747742, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.67095947265625, "learning_rate": 1e-06, "loss": 0.3425, "mean_token_accuracy": 0.8919684290885925, "num_tokens": 480970295.0, "step": 12609 }, { "epoch": 1.6041216130263325, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.632513999938965, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8709875345230103, "num_tokens": 481011213.0, "step": 12610 }, { "epoch": 1.604248823304923, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.674398422241211, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.864084005355835, "num_tokens": 481045273.0, "step": 12611 }, { "epoch": 1.6043760335835135, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.651219367980957, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8507534265518188, "num_tokens": 481084567.0, "step": 12612 }, { "epoch": 1.604503243862104, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.725165367126465, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8710575103759766, "num_tokens": 481119455.0, "step": 12613 }, { "epoch": 1.6046304541406946, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.708121299743652, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8550946712493896, "num_tokens": 481154718.0, "step": 12614 }, { "epoch": 1.6047576644192851, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.693802833557129, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8639130592346191, "num_tokens": 481193716.0, "step": 12615 }, { "epoch": 1.6048848746978757, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.669848442077637, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8460081219673157, "num_tokens": 481231378.0, "step": 12616 }, { "epoch": 1.6050120849764662, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.70852279663086, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8680567741394043, "num_tokens": 481267907.0, "step": 12617 }, { "epoch": 1.6051392952550567, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.641788482666016, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8709288239479065, "num_tokens": 481309899.0, "step": 12618 }, { "epoch": 1.6052665055336472, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.670869827270508, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8795446753501892, "num_tokens": 481348845.0, "step": 12619 }, { "epoch": 1.6053937158122378, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.682405471801758, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8601658344268799, "num_tokens": 481384999.0, "step": 12620 }, { "epoch": 1.6055209260908283, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.680990219116211, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8720211386680603, "num_tokens": 481417287.0, "step": 12621 }, { "epoch": 1.6056481363694186, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.701761245727539, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8731064796447754, "num_tokens": 481455943.0, "step": 12622 }, { "epoch": 1.6057753466480091, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.681200981140137, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8840265274047852, "num_tokens": 481492017.0, "step": 12623 }, { "epoch": 1.6059025569265997, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.696800231933594, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8627347946166992, "num_tokens": 481535115.0, "step": 12624 }, { "epoch": 1.6060297672051902, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.638744354248047, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8542962670326233, "num_tokens": 481580048.0, "step": 12625 }, { "epoch": 1.6061569774837807, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.699675559997559, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8804540038108826, "num_tokens": 481616748.0, "step": 12626 }, { "epoch": 1.6062841877623713, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.722442626953125, "learning_rate": 1e-06, "loss": 0.4915, "mean_token_accuracy": 0.8401431441307068, "num_tokens": 481659123.0, "step": 12627 }, { "epoch": 1.6064113980409616, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.705436706542969, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8598428964614868, "num_tokens": 481704791.0, "step": 12628 }, { "epoch": 1.606538608319552, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.713357925415039, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8567851185798645, "num_tokens": 481742913.0, "step": 12629 }, { "epoch": 1.6066658185981426, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.686199188232422, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.857010006904602, "num_tokens": 481783549.0, "step": 12630 }, { "epoch": 1.6067930288767331, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.668148040771484, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8748264312744141, "num_tokens": 481818486.0, "step": 12631 }, { "epoch": 1.6069202391553237, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.664411544799805, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8695118427276611, "num_tokens": 481849969.0, "step": 12632 }, { "epoch": 1.6070474494339142, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.708460807800293, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8532775640487671, "num_tokens": 481886108.0, "step": 12633 }, { "epoch": 1.6071746597125047, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.694900512695312, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8801396489143372, "num_tokens": 481919389.0, "step": 12634 }, { "epoch": 1.6073018699910953, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.630782127380371, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8824401497840881, "num_tokens": 481958531.0, "step": 12635 }, { "epoch": 1.6074290802696858, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.67973518371582, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8651440739631653, "num_tokens": 481999864.0, "step": 12636 }, { "epoch": 1.6075562905482763, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.64645767211914, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8919601440429688, "num_tokens": 482033890.0, "step": 12637 }, { "epoch": 1.6076835008268668, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.645782470703125, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8746281862258911, "num_tokens": 482071186.0, "step": 12638 }, { "epoch": 1.6078107111054574, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.67937183380127, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8673981428146362, "num_tokens": 482107218.0, "step": 12639 }, { "epoch": 1.607937921384048, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.612972259521484, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8657103776931763, "num_tokens": 482146101.0, "step": 12640 }, { "epoch": 1.6080651316626384, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.656927108764648, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8823326826095581, "num_tokens": 482178531.0, "step": 12641 }, { "epoch": 1.608192341941229, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.701970100402832, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8776743412017822, "num_tokens": 482217019.0, "step": 12642 }, { "epoch": 1.6083195522198195, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.639678955078125, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8705505728721619, "num_tokens": 482250547.0, "step": 12643 }, { "epoch": 1.60844676249841, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.748945236206055, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8657171130180359, "num_tokens": 482287009.0, "step": 12644 }, { "epoch": 1.6085739727770005, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.6470365524292, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8561549186706543, "num_tokens": 482327373.0, "step": 12645 }, { "epoch": 1.6087011830555908, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.594669342041016, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8577379584312439, "num_tokens": 482363409.0, "step": 12646 }, { "epoch": 1.6088283933341814, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.72310733795166, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8600424528121948, "num_tokens": 482396499.0, "step": 12647 }, { "epoch": 1.608955603612772, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.590292930603027, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8707671165466309, "num_tokens": 482435059.0, "step": 12648 }, { "epoch": 1.6090828138913624, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.656436920166016, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8622395396232605, "num_tokens": 482472793.0, "step": 12649 }, { "epoch": 1.609210024169953, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.648609161376953, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8716591596603394, "num_tokens": 482508857.0, "step": 12650 }, { "epoch": 1.6093372344485435, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 13.198454856872559, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.852882981300354, "num_tokens": 482548526.0, "step": 12651 }, { "epoch": 1.6094644447271338, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.60049057006836, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8619015216827393, "num_tokens": 482586461.0, "step": 12652 }, { "epoch": 1.6095916550057243, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.64285659790039, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8618152141571045, "num_tokens": 482624642.0, "step": 12653 }, { "epoch": 1.6097188652843148, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.644083023071289, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8681451082229614, "num_tokens": 482664096.0, "step": 12654 }, { "epoch": 1.6098460755629054, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.705602645874023, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8795474767684937, "num_tokens": 482701230.0, "step": 12655 }, { "epoch": 1.609973285841496, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.581710815429688, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8689841628074646, "num_tokens": 482741611.0, "step": 12656 }, { "epoch": 1.6101004961200864, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.684560775756836, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8791720271110535, "num_tokens": 482779482.0, "step": 12657 }, { "epoch": 1.610227706398677, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.620315551757812, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8773232102394104, "num_tokens": 482816279.0, "step": 12658 }, { "epoch": 1.6103549166772675, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.652253150939941, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.861319899559021, "num_tokens": 482859941.0, "step": 12659 }, { "epoch": 1.610482126955858, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.695297241210938, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8738964200019836, "num_tokens": 482894348.0, "step": 12660 }, { "epoch": 1.6106093372344485, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.682506561279297, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8634000420570374, "num_tokens": 482930882.0, "step": 12661 }, { "epoch": 1.610736547513039, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.721158027648926, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8521820306777954, "num_tokens": 482969515.0, "step": 12662 }, { "epoch": 1.6108637577916296, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.675700187683105, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8667457103729248, "num_tokens": 483008074.0, "step": 12663 }, { "epoch": 1.6109909680702201, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.718157768249512, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8615747094154358, "num_tokens": 483038920.0, "step": 12664 }, { "epoch": 1.6111181783488107, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.620909690856934, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8803372979164124, "num_tokens": 483075559.0, "step": 12665 }, { "epoch": 1.6112453886274012, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.683204650878906, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8718876242637634, "num_tokens": 483114378.0, "step": 12666 }, { "epoch": 1.6113725989059917, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.70012092590332, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8730994462966919, "num_tokens": 483147911.0, "step": 12667 }, { "epoch": 1.6114998091845822, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.662577629089355, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.88371741771698, "num_tokens": 483188275.0, "step": 12668 }, { "epoch": 1.6116270194631728, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.681402206420898, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8797404170036316, "num_tokens": 483230804.0, "step": 12669 }, { "epoch": 1.6117542297417633, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.69428825378418, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8715617060661316, "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": 12.689661979675293, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.8579993844032288, "num_tokens": 483300894.0, "step": 12671 }, { "epoch": 1.6120086502989441, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.700952529907227, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8751416206359863, "num_tokens": 483337377.0, "step": 12672 }, { "epoch": 1.6121358605775347, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.673295021057129, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8682986497879028, "num_tokens": 483376572.0, "step": 12673 }, { "epoch": 1.6122630708561252, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.66041374206543, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8572427034378052, "num_tokens": 483415584.0, "step": 12674 }, { "epoch": 1.6123902811347157, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.660390853881836, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8700991272926331, "num_tokens": 483459584.0, "step": 12675 }, { "epoch": 1.6125174914133062, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.597718238830566, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8634780645370483, "num_tokens": 483504260.0, "step": 12676 }, { "epoch": 1.6126447016918966, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.617002487182617, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8715119957923889, "num_tokens": 483547271.0, "step": 12677 }, { "epoch": 1.612771911970487, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.65279483795166, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8710069060325623, "num_tokens": 483582255.0, "step": 12678 }, { "epoch": 1.6128991222490776, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.67556095123291, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8647679090499878, "num_tokens": 483622290.0, "step": 12679 }, { "epoch": 1.6130263325276681, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.674797058105469, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8814629316329956, "num_tokens": 483657253.0, "step": 12680 }, { "epoch": 1.6131535428062587, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.670997619628906, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8730200529098511, "num_tokens": 483693062.0, "step": 12681 }, { "epoch": 1.6132807530848492, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.655961990356445, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.864040732383728, "num_tokens": 483735845.0, "step": 12682 }, { "epoch": 1.6134079633634397, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.680665969848633, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8731138706207275, "num_tokens": 483783833.0, "step": 12683 }, { "epoch": 1.6135351736420303, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.669889450073242, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8682003021240234, "num_tokens": 483820175.0, "step": 12684 }, { "epoch": 1.6136623839206208, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.697989463806152, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8809350728988647, "num_tokens": 483863454.0, "step": 12685 }, { "epoch": 1.6137895941992113, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.66611385345459, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.889751672744751, "num_tokens": 483900657.0, "step": 12686 }, { "epoch": 1.6139168044778018, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.669907569885254, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8632987141609192, "num_tokens": 483935912.0, "step": 12687 }, { "epoch": 1.6140440147563924, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.730658531188965, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8692734241485596, "num_tokens": 483968594.0, "step": 12688 }, { "epoch": 1.614171225034983, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.592690467834473, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8786766529083252, "num_tokens": 484010866.0, "step": 12689 }, { "epoch": 1.6142984353135734, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.673813819885254, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8636903762817383, "num_tokens": 484051355.0, "step": 12690 }, { "epoch": 1.614425645592164, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.628929138183594, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8794363737106323, "num_tokens": 484087049.0, "step": 12691 }, { "epoch": 1.6145528558707545, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.706633567810059, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8583517074584961, "num_tokens": 484124914.0, "step": 12692 }, { "epoch": 1.614680066149345, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.677450180053711, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8861665725708008, "num_tokens": 484154524.0, "step": 12693 }, { "epoch": 1.6148072764279355, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.735255241394043, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8682530522346497, "num_tokens": 484195091.0, "step": 12694 }, { "epoch": 1.6149344867065258, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.622587203979492, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8719479441642761, "num_tokens": 484236787.0, "step": 12695 }, { "epoch": 1.6150616969851164, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.69249153137207, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8647462129592896, "num_tokens": 484281748.0, "step": 12696 }, { "epoch": 1.615188907263707, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.690478324890137, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8751596212387085, "num_tokens": 484321181.0, "step": 12697 }, { "epoch": 1.6153161175422974, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.687454223632812, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.866880476474762, "num_tokens": 484360476.0, "step": 12698 }, { "epoch": 1.615443327820888, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.670672416687012, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8713845610618591, "num_tokens": 484395065.0, "step": 12699 }, { "epoch": 1.6155705380994785, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.665482521057129, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8718600273132324, "num_tokens": 484439309.0, "step": 12700 }, { "epoch": 1.6156977483780688, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.704577445983887, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8703489899635315, "num_tokens": 484473018.0, "step": 12701 }, { "epoch": 1.6158249586566593, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.70242691040039, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8764027953147888, "num_tokens": 484518306.0, "step": 12702 }, { "epoch": 1.6159521689352498, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.700357437133789, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8566249012947083, "num_tokens": 484556348.0, "step": 12703 }, { "epoch": 1.6160793792138404, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.659934043884277, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8807128667831421, "num_tokens": 484593136.0, "step": 12704 }, { "epoch": 1.616206589492431, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.70671272277832, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8793973326683044, "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": 12.689051628112793, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8754763603210449, "num_tokens": 484666703.0, "step": 12706 }, { "epoch": 1.616461010049612, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.700254440307617, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8794600963592529, "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": 12.660897254943848, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8843910694122314, "num_tokens": 484741830.0, "step": 12708 }, { "epoch": 1.616715430606793, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.745269775390625, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.868618369102478, "num_tokens": 484780140.0, "step": 12709 }, { "epoch": 1.6168426408853835, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.693775177001953, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8683180809020996, "num_tokens": 484813856.0, "step": 12710 }, { "epoch": 1.616969851163974, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.659913063049316, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8785789608955383, "num_tokens": 484851383.0, "step": 12711 }, { "epoch": 1.6170970614425646, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.68641185760498, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8787869811058044, "num_tokens": 484884361.0, "step": 12712 }, { "epoch": 1.6172242717211551, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.689632415771484, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.848479688167572, "num_tokens": 484922106.0, "step": 12713 }, { "epoch": 1.6173514819997457, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.720818519592285, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8705289959907532, "num_tokens": 484963100.0, "step": 12714 }, { "epoch": 1.6174786922783362, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.638986587524414, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8813358545303345, "num_tokens": 485006896.0, "step": 12715 }, { "epoch": 1.6176059025569267, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.639153480529785, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8709507584571838, "num_tokens": 485045968.0, "step": 12716 }, { "epoch": 1.6177331128355172, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.659476280212402, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8671083450317383, "num_tokens": 485084736.0, "step": 12717 }, { "epoch": 1.6178603231141078, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.66850471496582, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8782140612602234, "num_tokens": 485123130.0, "step": 12718 }, { "epoch": 1.6179875333926983, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.699620246887207, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8664532899856567, "num_tokens": 485158072.0, "step": 12719 }, { "epoch": 1.6181147436712886, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.64975643157959, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8783277273178101, "num_tokens": 485196783.0, "step": 12720 }, { "epoch": 1.6182419539498791, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.732563018798828, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8762986660003662, "num_tokens": 485231433.0, "step": 12721 }, { "epoch": 1.6183691642284697, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.700216293334961, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8679022192955017, "num_tokens": 485269016.0, "step": 12722 }, { "epoch": 1.6184963745070602, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.66159725189209, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8600971102714539, "num_tokens": 485308293.0, "step": 12723 }, { "epoch": 1.6186235847856507, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.744505882263184, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8794471025466919, "num_tokens": 485340758.0, "step": 12724 }, { "epoch": 1.6187507950642412, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.706283569335938, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8851116299629211, "num_tokens": 485384047.0, "step": 12725 }, { "epoch": 1.6188780053428315, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.752302169799805, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8738123774528503, "num_tokens": 485416909.0, "step": 12726 }, { "epoch": 1.619005215621422, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.751461029052734, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8705480098724365, "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": 12.71320915222168, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8805923461914062, "num_tokens": 485491171.0, "step": 12728 }, { "epoch": 1.6192596361786031, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.669676780700684, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8656609058380127, "num_tokens": 485525511.0, "step": 12729 }, { "epoch": 1.6193868464571937, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.702725410461426, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8673042058944702, "num_tokens": 485562517.0, "step": 12730 }, { "epoch": 1.6195140567357842, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.703988075256348, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8718447685241699, "num_tokens": 485601058.0, "step": 12731 }, { "epoch": 1.6196412670143747, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.739720344543457, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8836519718170166, "num_tokens": 485634737.0, "step": 12732 }, { "epoch": 1.6197684772929652, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.67331314086914, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8744391798973083, "num_tokens": 485671300.0, "step": 12733 }, { "epoch": 1.6198956875715558, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.716889381408691, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8785306811332703, "num_tokens": 485708712.0, "step": 12734 }, { "epoch": 1.6200228978501463, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.688726425170898, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8833187222480774, "num_tokens": 485746820.0, "step": 12735 }, { "epoch": 1.6201501081287368, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.670547485351562, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8707976341247559, "num_tokens": 485785921.0, "step": 12736 }, { "epoch": 1.6202773184073274, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.686928749084473, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.871000349521637, "num_tokens": 485825766.0, "step": 12737 }, { "epoch": 1.6204045286859179, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.693896293640137, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8648182153701782, "num_tokens": 485862288.0, "step": 12738 }, { "epoch": 1.6205317389645084, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.632747650146484, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8639130592346191, "num_tokens": 485903582.0, "step": 12739 }, { "epoch": 1.620658949243099, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.66318130493164, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8676624298095703, "num_tokens": 485950931.0, "step": 12740 }, { "epoch": 1.6207861595216895, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.699546813964844, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8730601668357849, "num_tokens": 485984185.0, "step": 12741 }, { "epoch": 1.62091336980028, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.668313026428223, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8758730888366699, "num_tokens": 486018796.0, "step": 12742 }, { "epoch": 1.6210405800788705, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.730635643005371, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8577367067337036, "num_tokens": 486055328.0, "step": 12743 }, { "epoch": 1.6211677903574608, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.676139831542969, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8650993704795837, "num_tokens": 486092340.0, "step": 12744 }, { "epoch": 1.6212950006360514, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.699954986572266, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8576374650001526, "num_tokens": 486123195.0, "step": 12745 }, { "epoch": 1.621422210914642, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.655477523803711, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.88051438331604, "num_tokens": 486167145.0, "step": 12746 }, { "epoch": 1.6215494211932324, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.669271469116211, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8623486161231995, "num_tokens": 486207925.0, "step": 12747 }, { "epoch": 1.621676631471823, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.673749923706055, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8749899864196777, "num_tokens": 486245446.0, "step": 12748 }, { "epoch": 1.6218038417504135, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.67821216583252, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.870725691318512, "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": 12.631644248962402, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8589872717857361, "num_tokens": 486323926.0, "step": 12750 }, { "epoch": 1.6220582623075943, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.67696475982666, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8645243048667908, "num_tokens": 486360016.0, "step": 12751 }, { "epoch": 1.6221854725861848, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.666147232055664, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8732338547706604, "num_tokens": 486394994.0, "step": 12752 }, { "epoch": 1.6223126828647754, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.641141891479492, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8699190616607666, "num_tokens": 486436909.0, "step": 12753 }, { "epoch": 1.622439893143366, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.679250717163086, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8772054314613342, "num_tokens": 486473201.0, "step": 12754 }, { "epoch": 1.6225671034219564, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.705814361572266, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8520122170448303, "num_tokens": 486510733.0, "step": 12755 }, { "epoch": 1.622694313700547, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.651724815368652, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8855051398277283, "num_tokens": 486546666.0, "step": 12756 }, { "epoch": 1.6228215239791375, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.734949111938477, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8517553210258484, "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": 12.649879455566406, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8755398988723755, "num_tokens": 486623028.0, "step": 12758 }, { "epoch": 1.6230759445363185, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.735532760620117, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8654496669769287, "num_tokens": 486655906.0, "step": 12759 }, { "epoch": 1.623203154814909, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.720528602600098, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8679330348968506, "num_tokens": 486689461.0, "step": 12760 }, { "epoch": 1.6233303650934996, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.689231872558594, "learning_rate": 1e-06, "loss": 0.4588, "mean_token_accuracy": 0.8552199006080627, "num_tokens": 486728373.0, "step": 12761 }, { "epoch": 1.6234575753720901, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.686847686767578, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8793473243713379, "num_tokens": 486769955.0, "step": 12762 }, { "epoch": 1.6235847856506807, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.737483024597168, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8747562170028687, "num_tokens": 486806468.0, "step": 12763 }, { "epoch": 1.6237119959292712, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.702226638793945, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8573501706123352, "num_tokens": 486846526.0, "step": 12764 }, { "epoch": 1.6238392062078617, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.705818176269531, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8879753947257996, "num_tokens": 486886323.0, "step": 12765 }, { "epoch": 1.6239664164864522, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.706835746765137, "learning_rate": 1e-06, "loss": 0.4545, "mean_token_accuracy": 0.8524539470672607, "num_tokens": 486918099.0, "step": 12766 }, { "epoch": 1.6240936267650428, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.727656364440918, "learning_rate": 1e-06, "loss": 0.3595, "mean_token_accuracy": 0.885211169719696, "num_tokens": 486953789.0, "step": 12767 }, { "epoch": 1.6242208370436333, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.693388938903809, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8694628477096558, "num_tokens": 486996165.0, "step": 12768 }, { "epoch": 1.6243480473222236, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.709917068481445, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8814164400100708, "num_tokens": 487028249.0, "step": 12769 }, { "epoch": 1.6244752576008141, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.727652549743652, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.870825469493866, "num_tokens": 487070099.0, "step": 12770 }, { "epoch": 1.6246024678794047, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.717052459716797, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8747073411941528, "num_tokens": 487107117.0, "step": 12771 }, { "epoch": 1.6247296781579952, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.712882995605469, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8587894439697266, "num_tokens": 487146024.0, "step": 12772 }, { "epoch": 1.6248568884365857, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.722468376159668, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8585047125816345, "num_tokens": 487189864.0, "step": 12773 }, { "epoch": 1.6249840987151762, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.643375396728516, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8728737831115723, "num_tokens": 487230307.0, "step": 12774 }, { "epoch": 1.6251113089937665, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.726097106933594, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8742239475250244, "num_tokens": 487261387.0, "step": 12775 }, { "epoch": 1.625238519272357, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.692657470703125, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8595834970474243, "num_tokens": 487298611.0, "step": 12776 }, { "epoch": 1.6253657295509476, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.651392936706543, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8703492879867554, "num_tokens": 487338172.0, "step": 12777 }, { "epoch": 1.6254929398295381, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.699974060058594, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8717828989028931, "num_tokens": 487375126.0, "step": 12778 }, { "epoch": 1.6256201501081287, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.647215843200684, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.881688117980957, "num_tokens": 487416539.0, "step": 12779 }, { "epoch": 1.6257473603867192, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.695249557495117, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8726996779441833, "num_tokens": 487448429.0, "step": 12780 }, { "epoch": 1.6258745706653097, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.701640129089355, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8699519038200378, "num_tokens": 487486054.0, "step": 12781 }, { "epoch": 1.6260017809439002, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.665190696716309, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8759635090827942, "num_tokens": 487523771.0, "step": 12782 }, { "epoch": 1.6261289912224908, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.674714088439941, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8763087391853333, "num_tokens": 487561334.0, "step": 12783 }, { "epoch": 1.6262562015010813, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.700332641601562, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.871403157711029, "num_tokens": 487606382.0, "step": 12784 }, { "epoch": 1.6263834117796718, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.663531303405762, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8659182786941528, "num_tokens": 487647640.0, "step": 12785 }, { "epoch": 1.6265106220582624, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.705096244812012, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8674481511116028, "num_tokens": 487682604.0, "step": 12786 }, { "epoch": 1.6266378323368529, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.700798988342285, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8744596242904663, "num_tokens": 487718271.0, "step": 12787 }, { "epoch": 1.6267650426154434, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.762791633605957, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8583874702453613, "num_tokens": 487752872.0, "step": 12788 }, { "epoch": 1.626892252894034, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.6360445022583, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8611321449279785, "num_tokens": 487798309.0, "step": 12789 }, { "epoch": 1.6270194631726245, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.679220199584961, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8597302436828613, "num_tokens": 487834052.0, "step": 12790 }, { "epoch": 1.627146673451215, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.74632740020752, "learning_rate": 1e-06, "loss": 0.4567, "mean_token_accuracy": 0.8587985038757324, "num_tokens": 487868662.0, "step": 12791 }, { "epoch": 1.6272738837298055, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.726568222045898, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.868739128112793, "num_tokens": 487907425.0, "step": 12792 }, { "epoch": 1.6274010940083958, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.727132797241211, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8791636228561401, "num_tokens": 487942979.0, "step": 12793 }, { "epoch": 1.6275283042869864, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.70828914642334, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8754117488861084, "num_tokens": 487982402.0, "step": 12794 }, { "epoch": 1.6276555145655769, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.726637840270996, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8774656057357788, "num_tokens": 488020071.0, "step": 12795 }, { "epoch": 1.6277827248441674, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.74841594696045, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.857538104057312, "num_tokens": 488057457.0, "step": 12796 }, { "epoch": 1.627909935122758, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.712544441223145, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8675135374069214, "num_tokens": 488094002.0, "step": 12797 }, { "epoch": 1.6280371454013485, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.697665214538574, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8559887409210205, "num_tokens": 488134995.0, "step": 12798 }, { "epoch": 1.6281643556799388, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.717327117919922, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8738028407096863, "num_tokens": 488169942.0, "step": 12799 }, { "epoch": 1.6282915659585293, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.710704803466797, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8719543218612671, "num_tokens": 488210852.0, "step": 12800 }, { "epoch": 1.6284187762371198, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.646456718444824, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8740037083625793, "num_tokens": 488253067.0, "step": 12801 }, { "epoch": 1.6285459865157104, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.75649642944336, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8778844475746155, "num_tokens": 488293498.0, "step": 12802 }, { "epoch": 1.628673196794301, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.699824333190918, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.871771514415741, "num_tokens": 488333931.0, "step": 12803 }, { "epoch": 1.6288004070728914, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.703795433044434, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8748847842216492, "num_tokens": 488371724.0, "step": 12804 }, { "epoch": 1.628927617351482, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.67485523223877, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8653154373168945, "num_tokens": 488415171.0, "step": 12805 }, { "epoch": 1.6290548276300725, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.682387351989746, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.864064633846283, "num_tokens": 488451672.0, "step": 12806 }, { "epoch": 1.629182037908663, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.69388484954834, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8850263357162476, "num_tokens": 488492657.0, "step": 12807 }, { "epoch": 1.6293092481872535, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.659639358520508, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8778471350669861, "num_tokens": 488530681.0, "step": 12808 }, { "epoch": 1.629436458465844, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.692418098449707, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8799763321876526, "num_tokens": 488560542.0, "step": 12809 }, { "epoch": 1.6295636687444346, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.715481758117676, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8721087574958801, "num_tokens": 488597491.0, "step": 12810 }, { "epoch": 1.6296908790230251, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.716200828552246, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8778852224349976, "num_tokens": 488634556.0, "step": 12811 }, { "epoch": 1.6298180893016156, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.69283676147461, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8671707510948181, "num_tokens": 488668174.0, "step": 12812 }, { "epoch": 1.6299452995802062, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.707012176513672, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8727842569351196, "num_tokens": 488705994.0, "step": 12813 }, { "epoch": 1.6300725098587967, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.731672286987305, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8597894906997681, "num_tokens": 488746908.0, "step": 12814 }, { "epoch": 1.6301997201373872, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.633755683898926, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.870692789554596, "num_tokens": 488785607.0, "step": 12815 }, { "epoch": 1.6303269304159778, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.768857955932617, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.868660569190979, "num_tokens": 488823831.0, "step": 12816 }, { "epoch": 1.630454140694568, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.72453498840332, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.878082811832428, "num_tokens": 488859944.0, "step": 12817 }, { "epoch": 1.6305813509731586, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.701081275939941, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8564308881759644, "num_tokens": 488903764.0, "step": 12818 }, { "epoch": 1.6307085612517491, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.678211212158203, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8604235053062439, "num_tokens": 488946890.0, "step": 12819 }, { "epoch": 1.6308357715303397, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.795373916625977, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.867242157459259, "num_tokens": 488981901.0, "step": 12820 }, { "epoch": 1.6309629818089302, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.646882057189941, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8734674453735352, "num_tokens": 489024180.0, "step": 12821 }, { "epoch": 1.6310901920875207, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.7172212600708, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.87363600730896, "num_tokens": 489070971.0, "step": 12822 }, { "epoch": 1.6312174023661112, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.688385963439941, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8804974555969238, "num_tokens": 489111636.0, "step": 12823 }, { "epoch": 1.6313446126447015, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.739167213439941, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.868144154548645, "num_tokens": 489148677.0, "step": 12824 }, { "epoch": 1.631471822923292, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.714573860168457, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8588330149650574, "num_tokens": 489186222.0, "step": 12825 }, { "epoch": 1.6315990332018826, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.749146461486816, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8445303440093994, "num_tokens": 489223173.0, "step": 12826 }, { "epoch": 1.6317262434804731, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.754859924316406, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8705477714538574, "num_tokens": 489259236.0, "step": 12827 }, { "epoch": 1.6318534537590637, "ewc_loss": 0.030029296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 12.700300216674805, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.863193929195404, "num_tokens": 489298449.0, "step": 12828 }, { "epoch": 1.6319806640376542, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.75101375579834, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8498910665512085, "num_tokens": 489334800.0, "step": 12829 }, { "epoch": 1.6321078743162447, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.732515335083008, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8765099048614502, "num_tokens": 489378510.0, "step": 12830 }, { "epoch": 1.6322350845948352, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.670992851257324, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8592370748519897, "num_tokens": 489419267.0, "step": 12831 }, { "epoch": 1.6323622948734258, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.741479873657227, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.873893678188324, "num_tokens": 489454772.0, "step": 12832 }, { "epoch": 1.6324895051520163, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.713164329528809, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8649155497550964, "num_tokens": 489490146.0, "step": 12833 }, { "epoch": 1.6326167154306068, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.667255401611328, "learning_rate": 1e-06, "loss": 0.501, "mean_token_accuracy": 0.841008186340332, "num_tokens": 489532754.0, "step": 12834 }, { "epoch": 1.6327439257091974, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.744319915771484, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8669843673706055, "num_tokens": 489570378.0, "step": 12835 }, { "epoch": 1.6328711359877879, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.749065399169922, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8618726134300232, "num_tokens": 489604562.0, "step": 12836 }, { "epoch": 1.6329983462663784, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 12.665961265563965, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8771136999130249, "num_tokens": 489644060.0, "step": 12837 }, { "epoch": 1.633125556544969, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.759591102600098, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8716706037521362, "num_tokens": 489682693.0, "step": 12838 }, { "epoch": 1.6332527668235595, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.671889305114746, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8785840272903442, "num_tokens": 489720786.0, "step": 12839 }, { "epoch": 1.63337997710215, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.715904235839844, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8760473728179932, "num_tokens": 489758773.0, "step": 12840 }, { "epoch": 1.6335071873807405, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.71151065826416, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.871256947517395, "num_tokens": 489793436.0, "step": 12841 }, { "epoch": 1.6336343976593308, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.662399291992188, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.8883987665176392, "num_tokens": 489828634.0, "step": 12842 }, { "epoch": 1.6337616079379214, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.750420570373535, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8732699751853943, "num_tokens": 489866335.0, "step": 12843 }, { "epoch": 1.6338888182165119, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.73388385772705, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8778872489929199, "num_tokens": 489903059.0, "step": 12844 }, { "epoch": 1.6340160284951024, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.688621520996094, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8771321177482605, "num_tokens": 489940186.0, "step": 12845 }, { "epoch": 1.634143238773693, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.682785987854004, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8818301558494568, "num_tokens": 489984732.0, "step": 12846 }, { "epoch": 1.6342704490522835, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.73145866394043, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8712021112442017, "num_tokens": 490016699.0, "step": 12847 }, { "epoch": 1.6343976593308738, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.713111877441406, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8626247644424438, "num_tokens": 490051351.0, "step": 12848 }, { "epoch": 1.6345248696094643, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.702289581298828, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8794960975646973, "num_tokens": 490089668.0, "step": 12849 }, { "epoch": 1.6346520798880548, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.686457633972168, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8716429471969604, "num_tokens": 490129651.0, "step": 12850 }, { "epoch": 1.6347792901666454, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.718711853027344, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8633358478546143, "num_tokens": 490168441.0, "step": 12851 }, { "epoch": 1.6349065004452359, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.726980209350586, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8804774284362793, "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": 12.70296859741211, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.88517165184021, "num_tokens": 490244178.0, "step": 12853 }, { "epoch": 1.635160921002417, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.757887840270996, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8661776781082153, "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": 12.691741943359375, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8718137741088867, "num_tokens": 490314892.0, "step": 12855 }, { "epoch": 1.635415341559598, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.744925498962402, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8601416349411011, "num_tokens": 490352600.0, "step": 12856 }, { "epoch": 1.6355425518381885, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.72515869140625, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8828315138816833, "num_tokens": 490384487.0, "step": 12857 }, { "epoch": 1.635669762116779, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.66985034942627, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8504919409751892, "num_tokens": 490422199.0, "step": 12858 }, { "epoch": 1.6357969723953696, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.73288345336914, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8632466793060303, "num_tokens": 490458058.0, "step": 12859 }, { "epoch": 1.6359241826739601, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.703609466552734, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8814351558685303, "num_tokens": 490498447.0, "step": 12860 }, { "epoch": 1.6360513929525506, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.66832447052002, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.857276201248169, "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": 12.695341110229492, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.874221920967102, "num_tokens": 490575752.0, "step": 12862 }, { "epoch": 1.6363058135097317, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.737902641296387, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8707587122917175, "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": 12.671160697937012, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8650107383728027, "num_tokens": 490653820.0, "step": 12864 }, { "epoch": 1.6365602340669128, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.684432029724121, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8556070923805237, "num_tokens": 490693332.0, "step": 12865 }, { "epoch": 1.636687444345503, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.731440544128418, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8688328266143799, "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": 12.694844245910645, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8733290433883667, "num_tokens": 490775505.0, "step": 12867 }, { "epoch": 1.6369418649026841, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.676677703857422, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8701401948928833, "num_tokens": 490817783.0, "step": 12868 }, { "epoch": 1.6370690751812746, "ewc_loss": 0.0302734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.731934547424316, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8643884658813477, "num_tokens": 490849464.0, "step": 12869 }, { "epoch": 1.6371962854598652, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.727361679077148, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8635537624359131, "num_tokens": 490891925.0, "step": 12870 }, { "epoch": 1.6373234957384557, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.732683181762695, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8640804290771484, "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": 12.725823402404785, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8615835309028625, "num_tokens": 490964141.0, "step": 12872 }, { "epoch": 1.6375779162956365, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.715890884399414, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.86576247215271, "num_tokens": 490997917.0, "step": 12873 }, { "epoch": 1.637705126574227, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.727019309997559, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8838500380516052, "num_tokens": 491029244.0, "step": 12874 }, { "epoch": 1.6378323368528176, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.72078800201416, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8692562580108643, "num_tokens": 491066591.0, "step": 12875 }, { "epoch": 1.6379595471314081, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.693281173706055, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8688186407089233, "num_tokens": 491106546.0, "step": 12876 }, { "epoch": 1.6380867574099987, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.700907707214355, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8629189729690552, "num_tokens": 491145512.0, "step": 12877 }, { "epoch": 1.6382139676885892, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.724233627319336, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8683863282203674, "num_tokens": 491185624.0, "step": 12878 }, { "epoch": 1.6383411779671797, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.713316917419434, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8640967607498169, "num_tokens": 491220982.0, "step": 12879 }, { "epoch": 1.6384683882457702, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.703211784362793, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8717095255851746, "num_tokens": 491262501.0, "step": 12880 }, { "epoch": 1.6385955985243608, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.707149505615234, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8816984295845032, "num_tokens": 491298078.0, "step": 12881 }, { "epoch": 1.6387228088029513, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.690674781799316, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8767269849777222, "num_tokens": 491336203.0, "step": 12882 }, { "epoch": 1.6388500190815418, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.688178062438965, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8586241006851196, "num_tokens": 491374723.0, "step": 12883 }, { "epoch": 1.6389772293601323, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.686856269836426, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8707455396652222, "num_tokens": 491410546.0, "step": 12884 }, { "epoch": 1.6391044396387229, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.71353530883789, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8660306930541992, "num_tokens": 491450642.0, "step": 12885 }, { "epoch": 1.6392316499173134, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.621611595153809, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8666472434997559, "num_tokens": 491489893.0, "step": 12886 }, { "epoch": 1.639358860195904, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.696783065795898, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8605066537857056, "num_tokens": 491535870.0, "step": 12887 }, { "epoch": 1.6394860704744945, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.685322761535645, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.868693470954895, "num_tokens": 491575988.0, "step": 12888 }, { "epoch": 1.639613280753085, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.632681846618652, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8716098070144653, "num_tokens": 491616822.0, "step": 12889 }, { "epoch": 1.6397404910316755, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.695050239562988, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8887369632720947, "num_tokens": 491654132.0, "step": 12890 }, { "epoch": 1.6398677013102658, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.708212852478027, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8568352460861206, "num_tokens": 491697215.0, "step": 12891 }, { "epoch": 1.6399949115888564, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.688182830810547, "learning_rate": 1e-06, "loss": 0.4769, "mean_token_accuracy": 0.8484658598899841, "num_tokens": 491731338.0, "step": 12892 }, { "epoch": 1.6401221218674469, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.732366561889648, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8688547611236572, "num_tokens": 491770808.0, "step": 12893 }, { "epoch": 1.6402493321460374, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.70004940032959, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.877995491027832, "num_tokens": 491804167.0, "step": 12894 }, { "epoch": 1.640376542424628, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.710227012634277, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8706488609313965, "num_tokens": 491840094.0, "step": 12895 }, { "epoch": 1.6405037527032185, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.722992897033691, "learning_rate": 1e-06, "loss": 0.4494, "mean_token_accuracy": 0.8559130430221558, "num_tokens": 491881643.0, "step": 12896 }, { "epoch": 1.6406309629818088, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.747373580932617, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8579014539718628, "num_tokens": 491920951.0, "step": 12897 }, { "epoch": 1.6407581732603993, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.661165237426758, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8653290867805481, "num_tokens": 491954599.0, "step": 12898 }, { "epoch": 1.6408853835389898, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.695110321044922, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8702213764190674, "num_tokens": 491995375.0, "step": 12899 }, { "epoch": 1.6410125938175804, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.684050559997559, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8646193742752075, "num_tokens": 492035297.0, "step": 12900 }, { "epoch": 1.6411398040961709, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.705400466918945, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8707613945007324, "num_tokens": 492071003.0, "step": 12901 }, { "epoch": 1.6412670143747614, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.739415168762207, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8699951171875, "num_tokens": 492106046.0, "step": 12902 }, { "epoch": 1.641394224653352, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.659819602966309, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8803018927574158, "num_tokens": 492147252.0, "step": 12903 }, { "epoch": 1.6415214349319425, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.7633695602417, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8608306646347046, "num_tokens": 492188185.0, "step": 12904 }, { "epoch": 1.641648645210533, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.693835258483887, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8751957416534424, "num_tokens": 492219630.0, "step": 12905 }, { "epoch": 1.6417758554891235, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.729393005371094, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8716322183609009, "num_tokens": 492251955.0, "step": 12906 }, { "epoch": 1.641903065767714, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.687479972839355, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8716162443161011, "num_tokens": 492293484.0, "step": 12907 }, { "epoch": 1.6420302760463046, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.72330093383789, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8545817732810974, "num_tokens": 492331169.0, "step": 12908 }, { "epoch": 1.642157486324895, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.644450187683105, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8768613338470459, "num_tokens": 492366071.0, "step": 12909 }, { "epoch": 1.6422846966034856, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.738101959228516, "learning_rate": 1e-06, "loss": 0.3427, "mean_token_accuracy": 0.8894830942153931, "num_tokens": 492396820.0, "step": 12910 }, { "epoch": 1.6424119068820762, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.693775177001953, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.876754641532898, "num_tokens": 492435282.0, "step": 12911 }, { "epoch": 1.6425391171606667, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.70323371887207, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8773312568664551, "num_tokens": 492470195.0, "step": 12912 }, { "epoch": 1.6426663274392572, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.705568313598633, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8643269538879395, "num_tokens": 492514265.0, "step": 12913 }, { "epoch": 1.6427935377178478, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.69919204711914, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8717622756958008, "num_tokens": 492550348.0, "step": 12914 }, { "epoch": 1.642920747996438, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.710515975952148, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8734402656555176, "num_tokens": 492590648.0, "step": 12915 }, { "epoch": 1.6430479582750286, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.674311637878418, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8718132972717285, "num_tokens": 492629523.0, "step": 12916 }, { "epoch": 1.6431751685536191, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.744356155395508, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8649482727050781, "num_tokens": 492668392.0, "step": 12917 }, { "epoch": 1.6433023788322096, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.738490104675293, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.859192430973053, "num_tokens": 492696073.0, "step": 12918 }, { "epoch": 1.6434295891108002, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.734795570373535, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.872485339641571, "num_tokens": 492728590.0, "step": 12919 }, { "epoch": 1.6435567993893907, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.72269058227539, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8620508909225464, "num_tokens": 492767290.0, "step": 12920 }, { "epoch": 1.6436840096679812, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.71733570098877, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8810417056083679, "num_tokens": 492801019.0, "step": 12921 }, { "epoch": 1.6438112199465715, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.718683242797852, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8689680099487305, "num_tokens": 492841220.0, "step": 12922 }, { "epoch": 1.643938430225162, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.675890922546387, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8590858578681946, "num_tokens": 492885076.0, "step": 12923 }, { "epoch": 1.6440656405037526, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.77564525604248, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8653554916381836, "num_tokens": 492918554.0, "step": 12924 }, { "epoch": 1.6441928507823431, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.710918426513672, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.8853511214256287, "num_tokens": 492961890.0, "step": 12925 }, { "epoch": 1.6443200610609336, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.689595222473145, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8795394897460938, "num_tokens": 492997768.0, "step": 12926 }, { "epoch": 1.6444472713395242, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.710603713989258, "learning_rate": 1e-06, "loss": 0.3347, "mean_token_accuracy": 0.8940935730934143, "num_tokens": 493037017.0, "step": 12927 }, { "epoch": 1.6445744816181147, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.719616889953613, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8802769780158997, "num_tokens": 493074579.0, "step": 12928 }, { "epoch": 1.6447016918967052, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.738314628601074, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8713973760604858, "num_tokens": 493111249.0, "step": 12929 }, { "epoch": 1.6448289021752958, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.681598663330078, "learning_rate": 1e-06, "loss": 0.4576, "mean_token_accuracy": 0.8545187711715698, "num_tokens": 493150723.0, "step": 12930 }, { "epoch": 1.6449561124538863, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.717079162597656, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8679262399673462, "num_tokens": 493189530.0, "step": 12931 }, { "epoch": 1.6450833227324768, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.705892562866211, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8768742084503174, "num_tokens": 493231369.0, "step": 12932 }, { "epoch": 1.6452105330110673, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.745519638061523, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.855951189994812, "num_tokens": 493278430.0, "step": 12933 }, { "epoch": 1.6453377432896579, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.728995323181152, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8748258352279663, "num_tokens": 493312258.0, "step": 12934 }, { "epoch": 1.6454649535682484, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.77949047088623, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.868767499923706, "num_tokens": 493347725.0, "step": 12935 }, { "epoch": 1.645592163846839, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.716991424560547, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8743107914924622, "num_tokens": 493388119.0, "step": 12936 }, { "epoch": 1.6457193741254295, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.73984432220459, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8714954853057861, "num_tokens": 493424607.0, "step": 12937 }, { "epoch": 1.64584658440402, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.793004035949707, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8471949696540833, "num_tokens": 493459546.0, "step": 12938 }, { "epoch": 1.6459737946826105, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.715876579284668, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8734434843063354, "num_tokens": 493494162.0, "step": 12939 }, { "epoch": 1.6461010049612008, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.782416343688965, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8681632280349731, "num_tokens": 493537523.0, "step": 12940 }, { "epoch": 1.6462282152397913, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.734384536743164, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8685988783836365, "num_tokens": 493580065.0, "step": 12941 }, { "epoch": 1.6463554255183819, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.730299949645996, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8763675689697266, "num_tokens": 493626717.0, "step": 12942 }, { "epoch": 1.6464826357969724, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.768065452575684, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8761776685714722, "num_tokens": 493666266.0, "step": 12943 }, { "epoch": 1.646609846075563, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.749638557434082, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8600257635116577, "num_tokens": 493703184.0, "step": 12944 }, { "epoch": 1.6467370563541535, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.723628044128418, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8846353888511658, "num_tokens": 493744061.0, "step": 12945 }, { "epoch": 1.6468642666327438, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.72739028930664, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8700382709503174, "num_tokens": 493783585.0, "step": 12946 }, { "epoch": 1.6469914769113343, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.749186515808105, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8765044212341309, "num_tokens": 493827781.0, "step": 12947 }, { "epoch": 1.6471186871899248, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.748235702514648, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8756544589996338, "num_tokens": 493864418.0, "step": 12948 }, { "epoch": 1.6472458974685154, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.760120391845703, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8692029714584351, "num_tokens": 493895226.0, "step": 12949 }, { "epoch": 1.6473731077471059, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.74321174621582, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8614615797996521, "num_tokens": 493935210.0, "step": 12950 }, { "epoch": 1.6475003180256964, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.75865650177002, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8701931834220886, "num_tokens": 493966368.0, "step": 12951 }, { "epoch": 1.647627528304287, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.761372566223145, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8638882637023926, "num_tokens": 494004304.0, "step": 12952 }, { "epoch": 1.6477547385828775, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.729546546936035, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.876856803894043, "num_tokens": 494044758.0, "step": 12953 }, { "epoch": 1.647881948861468, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.735514640808105, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8542882204055786, "num_tokens": 494080624.0, "step": 12954 }, { "epoch": 1.6480091591400585, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.80956745147705, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8765901327133179, "num_tokens": 494116215.0, "step": 12955 }, { "epoch": 1.648136369418649, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.735841751098633, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8679186105728149, "num_tokens": 494150317.0, "step": 12956 }, { "epoch": 1.6482635796972396, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.744959831237793, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8837834596633911, "num_tokens": 494187273.0, "step": 12957 }, { "epoch": 1.64839078997583, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.730677604675293, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8666369915008545, "num_tokens": 494226921.0, "step": 12958 }, { "epoch": 1.6485180002544206, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.759328842163086, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8773202896118164, "num_tokens": 494263141.0, "step": 12959 }, { "epoch": 1.6486452105330112, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.669217109680176, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.866396427154541, "num_tokens": 494307925.0, "step": 12960 }, { "epoch": 1.6487724208116017, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.73694896697998, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8638540506362915, "num_tokens": 494353428.0, "step": 12961 }, { "epoch": 1.6488996310901922, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.730804443359375, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8583937883377075, "num_tokens": 494383559.0, "step": 12962 }, { "epoch": 1.6490268413687827, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.707416534423828, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8735716342926025, "num_tokens": 494419564.0, "step": 12963 }, { "epoch": 1.649154051647373, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.752394676208496, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8729515075683594, "num_tokens": 494466886.0, "step": 12964 }, { "epoch": 1.6492812619259636, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.778529167175293, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8528096675872803, "num_tokens": 494506672.0, "step": 12965 }, { "epoch": 1.649408472204554, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.797886848449707, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8693958520889282, "num_tokens": 494541881.0, "step": 12966 }, { "epoch": 1.6495356824831446, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.704843521118164, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.874981164932251, "num_tokens": 494583546.0, "step": 12967 }, { "epoch": 1.6496628927617352, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.841943740844727, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8579773306846619, "num_tokens": 494616625.0, "step": 12968 }, { "epoch": 1.6497901030403257, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.756970405578613, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.883268415927887, "num_tokens": 494654590.0, "step": 12969 }, { "epoch": 1.6499173133189162, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.709135055541992, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8838282227516174, "num_tokens": 494692471.0, "step": 12970 }, { "epoch": 1.6500445235975065, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.766024589538574, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8620045781135559, "num_tokens": 494730347.0, "step": 12971 }, { "epoch": 1.650171733876097, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.742010116577148, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.858191192150116, "num_tokens": 494763753.0, "step": 12972 }, { "epoch": 1.6502989441546876, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.829440116882324, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8652397394180298, "num_tokens": 494800416.0, "step": 12973 }, { "epoch": 1.6504261544332781, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.716935157775879, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8687790632247925, "num_tokens": 494842733.0, "step": 12974 }, { "epoch": 1.6505533647118686, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.74808120727539, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8628648519515991, "num_tokens": 494878661.0, "step": 12975 }, { "epoch": 1.6506805749904592, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.755534172058105, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8628619909286499, "num_tokens": 494918142.0, "step": 12976 }, { "epoch": 1.6508077852690497, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.797441482543945, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8699406385421753, "num_tokens": 494956947.0, "step": 12977 }, { "epoch": 1.6509349955476402, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.793831825256348, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8653945326805115, "num_tokens": 494989850.0, "step": 12978 }, { "epoch": 1.6510622058262308, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.756220817565918, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8626285791397095, "num_tokens": 495026770.0, "step": 12979 }, { "epoch": 1.6511894161048213, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.778290748596191, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8809835910797119, "num_tokens": 495065009.0, "step": 12980 }, { "epoch": 1.6513166263834118, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.749527931213379, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8707985877990723, "num_tokens": 495096914.0, "step": 12981 }, { "epoch": 1.6514438366620023, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.81256103515625, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8741518259048462, "num_tokens": 495138348.0, "step": 12982 }, { "epoch": 1.6515710469405929, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.77055549621582, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8633878827095032, "num_tokens": 495177026.0, "step": 12983 }, { "epoch": 1.6516982572191834, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.741865158081055, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8717843294143677, "num_tokens": 495213850.0, "step": 12984 }, { "epoch": 1.651825467497774, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.75634765625, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8785263299942017, "num_tokens": 495257213.0, "step": 12985 }, { "epoch": 1.6519526777763645, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.766250610351562, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8604116439819336, "num_tokens": 495300018.0, "step": 12986 }, { "epoch": 1.652079888054955, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.744608879089355, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8676321506500244, "num_tokens": 495336471.0, "step": 12987 }, { "epoch": 1.6522070983335455, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.764141082763672, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8732370138168335, "num_tokens": 495374680.0, "step": 12988 }, { "epoch": 1.6523343086121358, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.737709999084473, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8727189898490906, "num_tokens": 495414096.0, "step": 12989 }, { "epoch": 1.6524615188907263, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.773249626159668, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8572224974632263, "num_tokens": 495452955.0, "step": 12990 }, { "epoch": 1.6525887291693169, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.757107734680176, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8673755526542664, "num_tokens": 495490091.0, "step": 12991 }, { "epoch": 1.6527159394479074, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.735448837280273, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8777212500572205, "num_tokens": 495524543.0, "step": 12992 }, { "epoch": 1.652843149726498, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.799870491027832, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8612239956855774, "num_tokens": 495562068.0, "step": 12993 }, { "epoch": 1.6529703600050885, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.707408905029297, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8774921894073486, "num_tokens": 495598355.0, "step": 12994 }, { "epoch": 1.6530975702836788, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.8246431350708, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8562501668930054, "num_tokens": 495634735.0, "step": 12995 }, { "epoch": 1.6532247805622693, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.77144718170166, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8702043890953064, "num_tokens": 495668804.0, "step": 12996 }, { "epoch": 1.6533519908408598, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.798192977905273, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.874313473701477, "num_tokens": 495707129.0, "step": 12997 }, { "epoch": 1.6534792011194503, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.745084762573242, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8492093086242676, "num_tokens": 495746501.0, "step": 12998 }, { "epoch": 1.6536064113980409, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.75296688079834, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8855453133583069, "num_tokens": 495789723.0, "step": 12999 }, { "epoch": 1.6537336216766314, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.771129608154297, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8599517345428467, "num_tokens": 495827073.0, "step": 13000 }, { "epoch": 1.653860831955222, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.779375076293945, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.85893315076828, "num_tokens": 495862362.0, "step": 13001 }, { "epoch": 1.6539880422338125, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.700203895568848, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.871559202671051, "num_tokens": 495905546.0, "step": 13002 }, { "epoch": 1.654115252512403, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.785735130310059, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.867661714553833, "num_tokens": 495942291.0, "step": 13003 }, { "epoch": 1.6542424627909935, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.779267311096191, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.859004020690918, "num_tokens": 495978631.0, "step": 13004 }, { "epoch": 1.654369673069584, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.81654167175293, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8633902072906494, "num_tokens": 496016626.0, "step": 13005 }, { "epoch": 1.6544968833481746, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.762497901916504, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8621547222137451, "num_tokens": 496051949.0, "step": 13006 }, { "epoch": 1.654624093626765, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.753117561340332, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8756093382835388, "num_tokens": 496092693.0, "step": 13007 }, { "epoch": 1.6547513039053556, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.7567720413208, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8635270595550537, "num_tokens": 496132853.0, "step": 13008 }, { "epoch": 1.6548785141839462, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.766525268554688, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8549953699111938, "num_tokens": 496168303.0, "step": 13009 }, { "epoch": 1.6550057244625367, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.795654296875, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8573956489562988, "num_tokens": 496206132.0, "step": 13010 }, { "epoch": 1.6551329347411272, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.720329284667969, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8633224368095398, "num_tokens": 496247542.0, "step": 13011 }, { "epoch": 1.6552601450197177, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.75277328491211, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8789238333702087, "num_tokens": 496285698.0, "step": 13012 }, { "epoch": 1.655387355298308, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.78358268737793, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8711700439453125, "num_tokens": 496322216.0, "step": 13013 }, { "epoch": 1.6555145655768986, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.79767894744873, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8675907850265503, "num_tokens": 496357687.0, "step": 13014 }, { "epoch": 1.655641775855489, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.706954956054688, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8809757232666016, "num_tokens": 496400875.0, "step": 13015 }, { "epoch": 1.6557689861340796, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.764692306518555, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8725605607032776, "num_tokens": 496432076.0, "step": 13016 }, { "epoch": 1.6558961964126702, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.776471138000488, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.866168737411499, "num_tokens": 496466852.0, "step": 13017 }, { "epoch": 1.6560234066912607, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.736130714416504, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8680235147476196, "num_tokens": 496508890.0, "step": 13018 }, { "epoch": 1.6561506169698512, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.723551750183105, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8849178552627563, "num_tokens": 496537964.0, "step": 13019 }, { "epoch": 1.6562778272484415, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.791205406188965, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8536256551742554, "num_tokens": 496571263.0, "step": 13020 }, { "epoch": 1.656405037527032, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.750606536865234, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8774319887161255, "num_tokens": 496610058.0, "step": 13021 }, { "epoch": 1.6565322478056226, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.787095069885254, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8632853031158447, "num_tokens": 496640542.0, "step": 13022 }, { "epoch": 1.656659458084213, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.76655387878418, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8678818345069885, "num_tokens": 496678030.0, "step": 13023 }, { "epoch": 1.6567866683628036, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.774149894714355, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8654533624649048, "num_tokens": 496714057.0, "step": 13024 }, { "epoch": 1.6569138786413942, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.745194435119629, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8843400478363037, "num_tokens": 496744501.0, "step": 13025 }, { "epoch": 1.6570410889199847, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.752540588378906, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.874122142791748, "num_tokens": 496780233.0, "step": 13026 }, { "epoch": 1.6571682991985752, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.763692855834961, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8706212043762207, "num_tokens": 496809645.0, "step": 13027 }, { "epoch": 1.6572955094771658, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.746395111083984, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8701512217521667, "num_tokens": 496854871.0, "step": 13028 }, { "epoch": 1.6574227197557563, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.75671672821045, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8851625919342041, "num_tokens": 496891958.0, "step": 13029 }, { "epoch": 1.6575499300343468, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.744941711425781, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8627780675888062, "num_tokens": 496928550.0, "step": 13030 }, { "epoch": 1.6576771403129373, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.678732872009277, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8678299188613892, "num_tokens": 496972009.0, "step": 13031 }, { "epoch": 1.6578043505915279, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.7495698928833, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8684975504875183, "num_tokens": 497013605.0, "step": 13032 }, { "epoch": 1.6579315608701184, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.747847557067871, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8706972599029541, "num_tokens": 497048712.0, "step": 13033 }, { "epoch": 1.658058771148709, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.693448066711426, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8824253082275391, "num_tokens": 497089161.0, "step": 13034 }, { "epoch": 1.6581859814272994, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.741935729980469, "learning_rate": 1e-06, "loss": 0.4937, "mean_token_accuracy": 0.8417890071868896, "num_tokens": 497127253.0, "step": 13035 }, { "epoch": 1.65831319170589, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.751534461975098, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8667325973510742, "num_tokens": 497167637.0, "step": 13036 }, { "epoch": 1.6584404019844805, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.725772857666016, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8695514798164368, "num_tokens": 497205663.0, "step": 13037 }, { "epoch": 1.6585676122630708, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.733684539794922, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8863016366958618, "num_tokens": 497242146.0, "step": 13038 }, { "epoch": 1.6586948225416613, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.728468894958496, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8735193014144897, "num_tokens": 497279339.0, "step": 13039 }, { "epoch": 1.6588220328202519, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.68442440032959, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8613410592079163, "num_tokens": 497318530.0, "step": 13040 }, { "epoch": 1.6589492430988424, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.759781837463379, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8631815910339355, "num_tokens": 497352943.0, "step": 13041 }, { "epoch": 1.659076453377433, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.723064422607422, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8640373945236206, "num_tokens": 497387729.0, "step": 13042 }, { "epoch": 1.6592036636560235, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.693194389343262, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8703565001487732, "num_tokens": 497427875.0, "step": 13043 }, { "epoch": 1.6593308739346138, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.716552734375, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8623977899551392, "num_tokens": 497469076.0, "step": 13044 }, { "epoch": 1.6594580842132043, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.71463394165039, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8707537651062012, "num_tokens": 497509556.0, "step": 13045 }, { "epoch": 1.6595852944917948, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.745089530944824, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8623090982437134, "num_tokens": 497548151.0, "step": 13046 }, { "epoch": 1.6597125047703853, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.709362030029297, "learning_rate": 1e-06, "loss": 0.4944, "mean_token_accuracy": 0.8436009883880615, "num_tokens": 497589088.0, "step": 13047 }, { "epoch": 1.6598397150489759, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.696542739868164, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8789581060409546, "num_tokens": 497632095.0, "step": 13048 }, { "epoch": 1.6599669253275664, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.774105072021484, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8809990882873535, "num_tokens": 497665148.0, "step": 13049 }, { "epoch": 1.660094135606157, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.758766174316406, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8631483316421509, "num_tokens": 497700134.0, "step": 13050 }, { "epoch": 1.6602213458847475, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.735801696777344, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8746209144592285, "num_tokens": 497743101.0, "step": 13051 }, { "epoch": 1.660348556163338, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.732015609741211, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8627241253852844, "num_tokens": 497775112.0, "step": 13052 }, { "epoch": 1.6604757664419285, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.767350196838379, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8803707957267761, "num_tokens": 497813320.0, "step": 13053 }, { "epoch": 1.660602976720519, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.681036949157715, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.865288496017456, "num_tokens": 497852780.0, "step": 13054 }, { "epoch": 1.6607301869991096, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.7757568359375, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.871319591999054, "num_tokens": 497886313.0, "step": 13055 }, { "epoch": 1.6608573972777, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.726637840270996, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.866561770439148, "num_tokens": 497924108.0, "step": 13056 }, { "epoch": 1.6609846075562906, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.724461555480957, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8589130640029907, "num_tokens": 497964656.0, "step": 13057 }, { "epoch": 1.6611118178348812, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.744232177734375, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8566439151763916, "num_tokens": 498005436.0, "step": 13058 }, { "epoch": 1.6612390281134717, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.701227188110352, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8655299544334412, "num_tokens": 498042237.0, "step": 13059 }, { "epoch": 1.6613662383920622, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.749748229980469, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8684537410736084, "num_tokens": 498083008.0, "step": 13060 }, { "epoch": 1.6614934486706527, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.729426383972168, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.846231997013092, "num_tokens": 498120550.0, "step": 13061 }, { "epoch": 1.661620658949243, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.793837547302246, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8671679496765137, "num_tokens": 498153033.0, "step": 13062 }, { "epoch": 1.6617478692278336, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.684347152709961, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8716081380844116, "num_tokens": 498190491.0, "step": 13063 }, { "epoch": 1.661875079506424, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.756059646606445, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8619983792304993, "num_tokens": 498235889.0, "step": 13064 }, { "epoch": 1.6620022897850146, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.715058326721191, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8662413954734802, "num_tokens": 498277191.0, "step": 13065 }, { "epoch": 1.6621295000636052, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.769515991210938, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8594720363616943, "num_tokens": 498318443.0, "step": 13066 }, { "epoch": 1.6622567103421957, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.796283721923828, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8567445874214172, "num_tokens": 498355955.0, "step": 13067 }, { "epoch": 1.662383920620786, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.752941131591797, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.869546115398407, "num_tokens": 498388588.0, "step": 13068 }, { "epoch": 1.6625111308993765, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.732584953308105, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8677244186401367, "num_tokens": 498430661.0, "step": 13069 }, { "epoch": 1.662638341177967, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.739828109741211, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8567817211151123, "num_tokens": 498470428.0, "step": 13070 }, { "epoch": 1.6627655514565576, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.723018646240234, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8684940338134766, "num_tokens": 498512031.0, "step": 13071 }, { "epoch": 1.662892761735148, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.783746719360352, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.8909868001937866, "num_tokens": 498548344.0, "step": 13072 }, { "epoch": 1.6630199720137386, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.69250774383545, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8823636770248413, "num_tokens": 498590115.0, "step": 13073 }, { "epoch": 1.6631471822923292, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.73896598815918, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8689911365509033, "num_tokens": 498634034.0, "step": 13074 }, { "epoch": 1.6632743925709197, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.779305458068848, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8614298701286316, "num_tokens": 498675086.0, "step": 13075 }, { "epoch": 1.6634016028495102, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.764998435974121, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8753839731216431, "num_tokens": 498711374.0, "step": 13076 }, { "epoch": 1.6635288131281007, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.722347259521484, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8734682202339172, "num_tokens": 498753050.0, "step": 13077 }, { "epoch": 1.6636560234066913, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.760353088378906, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8703484535217285, "num_tokens": 498790406.0, "step": 13078 }, { "epoch": 1.6637832336852818, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.776581764221191, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8678114414215088, "num_tokens": 498827712.0, "step": 13079 }, { "epoch": 1.6639104439638723, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.814773559570312, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8733747005462646, "num_tokens": 498862283.0, "step": 13080 }, { "epoch": 1.6640376542424629, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.716445922851562, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8702115416526794, "num_tokens": 498902686.0, "step": 13081 }, { "epoch": 1.6641648645210534, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.788039207458496, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8581567406654358, "num_tokens": 498942785.0, "step": 13082 }, { "epoch": 1.664292074799644, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.730876922607422, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8738667964935303, "num_tokens": 498982761.0, "step": 13083 }, { "epoch": 1.6644192850782344, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.827681541442871, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8505553603172302, "num_tokens": 499025762.0, "step": 13084 }, { "epoch": 1.664546495356825, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.795578002929688, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8603039979934692, "num_tokens": 499063394.0, "step": 13085 }, { "epoch": 1.6646737056354155, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.701501846313477, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8788818120956421, "num_tokens": 499105824.0, "step": 13086 }, { "epoch": 1.6648009159140058, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.841912269592285, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8774726986885071, "num_tokens": 499140935.0, "step": 13087 }, { "epoch": 1.6649281261925963, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.817753791809082, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8679991960525513, "num_tokens": 499179492.0, "step": 13088 }, { "epoch": 1.6650553364711869, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.776338577270508, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8795279264450073, "num_tokens": 499214716.0, "step": 13089 }, { "epoch": 1.6651825467497774, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.765585899353027, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8757851123809814, "num_tokens": 499250337.0, "step": 13090 }, { "epoch": 1.665309757028368, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.767179489135742, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8737077713012695, "num_tokens": 499283675.0, "step": 13091 }, { "epoch": 1.6654369673069584, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.801973342895508, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8743032813072205, "num_tokens": 499319153.0, "step": 13092 }, { "epoch": 1.6655641775855488, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.801042556762695, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.865807831287384, "num_tokens": 499356675.0, "step": 13093 }, { "epoch": 1.6656913878641393, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.71767520904541, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.883629560470581, "num_tokens": 499392129.0, "step": 13094 }, { "epoch": 1.6658185981427298, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.756171226501465, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8809382915496826, "num_tokens": 499424958.0, "step": 13095 }, { "epoch": 1.6659458084213203, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.745894432067871, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8778108954429626, "num_tokens": 499467769.0, "step": 13096 }, { "epoch": 1.6660730186999109, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.71920108795166, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8598732948303223, "num_tokens": 499513828.0, "step": 13097 }, { "epoch": 1.6662002289785014, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.738728523254395, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8749215006828308, "num_tokens": 499548219.0, "step": 13098 }, { "epoch": 1.666327439257092, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.823969841003418, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8523958325386047, "num_tokens": 499587158.0, "step": 13099 }, { "epoch": 1.6664546495356825, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.732707023620605, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8725978136062622, "num_tokens": 499625235.0, "step": 13100 }, { "epoch": 1.666581859814273, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.830761909484863, "learning_rate": 1e-06, "loss": 0.4586, "mean_token_accuracy": 0.8500816822052002, "num_tokens": 499662842.0, "step": 13101 }, { "epoch": 1.6667090700928635, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.851524353027344, "learning_rate": 1e-06, "loss": 0.4796, "mean_token_accuracy": 0.8463984727859497, "num_tokens": 499703464.0, "step": 13102 }, { "epoch": 1.666836280371454, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.793787002563477, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8682981133460999, "num_tokens": 499739160.0, "step": 13103 }, { "epoch": 1.6669634906500446, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.74903392791748, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.855782687664032, "num_tokens": 499775024.0, "step": 13104 }, { "epoch": 1.667090700928635, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.794299125671387, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8727425932884216, "num_tokens": 499816479.0, "step": 13105 }, { "epoch": 1.6672179112072256, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.788873672485352, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8702642917633057, "num_tokens": 499848790.0, "step": 13106 }, { "epoch": 1.6673451214858162, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.716010093688965, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8884477019309998, "num_tokens": 499890955.0, "step": 13107 }, { "epoch": 1.6674723317644067, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.758183479309082, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8782402873039246, "num_tokens": 499924813.0, "step": 13108 }, { "epoch": 1.6675995420429972, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.784466743469238, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8670137524604797, "num_tokens": 499960032.0, "step": 13109 }, { "epoch": 1.6677267523215877, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.73037052154541, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8835521936416626, "num_tokens": 499995868.0, "step": 13110 }, { "epoch": 1.667853962600178, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.75603199005127, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8739544153213501, "num_tokens": 500031662.0, "step": 13111 }, { "epoch": 1.6679811728787686, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.73603630065918, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8604926466941833, "num_tokens": 500070189.0, "step": 13112 }, { "epoch": 1.668108383157359, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.757880210876465, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8660540580749512, "num_tokens": 500109999.0, "step": 13113 }, { "epoch": 1.6682355934359496, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.751083374023438, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8534576296806335, "num_tokens": 500151473.0, "step": 13114 }, { "epoch": 1.6683628037145402, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.772994041442871, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8681588172912598, "num_tokens": 500187436.0, "step": 13115 }, { "epoch": 1.6684900139931307, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.783638000488281, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8695269227027893, "num_tokens": 500222433.0, "step": 13116 }, { "epoch": 1.668617224271721, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.826433181762695, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8691263198852539, "num_tokens": 500257390.0, "step": 13117 }, { "epoch": 1.6687444345503115, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.727879524230957, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8687206506729126, "num_tokens": 500301260.0, "step": 13118 }, { "epoch": 1.668871644828902, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.827539443969727, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8749027252197266, "num_tokens": 500343453.0, "step": 13119 }, { "epoch": 1.6689988551074926, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.794271469116211, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8765648007392883, "num_tokens": 500375133.0, "step": 13120 }, { "epoch": 1.669126065386083, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.779668807983398, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8651825189590454, "num_tokens": 500408540.0, "step": 13121 }, { "epoch": 1.6692532756646736, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.804757118225098, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.882173478603363, "num_tokens": 500446243.0, "step": 13122 }, { "epoch": 1.6693804859432642, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.80135726928711, "learning_rate": 1e-06, "loss": 0.496, "mean_token_accuracy": 0.8435735702514648, "num_tokens": 500484313.0, "step": 13123 }, { "epoch": 1.6695076962218547, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.719609260559082, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.881085991859436, "num_tokens": 500525566.0, "step": 13124 }, { "epoch": 1.6696349065004452, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.777050018310547, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8746389746665955, "num_tokens": 500566274.0, "step": 13125 }, { "epoch": 1.6697621167790357, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.814213752746582, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8628020286560059, "num_tokens": 500602769.0, "step": 13126 }, { "epoch": 1.6698893270576263, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.743897438049316, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8531405925750732, "num_tokens": 500642895.0, "step": 13127 }, { "epoch": 1.6700165373362168, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.795175552368164, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8681654930114746, "num_tokens": 500681461.0, "step": 13128 }, { "epoch": 1.6701437476148073, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.787721633911133, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.853594183921814, "num_tokens": 500723395.0, "step": 13129 }, { "epoch": 1.6702709578933979, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.771909713745117, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8708641529083252, "num_tokens": 500760571.0, "step": 13130 }, { "epoch": 1.6703981681719884, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.773686408996582, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.860392689704895, "num_tokens": 500797858.0, "step": 13131 }, { "epoch": 1.670525378450579, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.79511547088623, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.875613808631897, "num_tokens": 500840199.0, "step": 13132 }, { "epoch": 1.6706525887291694, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.827061653137207, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8640596866607666, "num_tokens": 500883290.0, "step": 13133 }, { "epoch": 1.67077979900776, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.794578552246094, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8729680776596069, "num_tokens": 500923072.0, "step": 13134 }, { "epoch": 1.6709070092863505, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.813898086547852, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8738032579421997, "num_tokens": 500958726.0, "step": 13135 }, { "epoch": 1.6710342195649408, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.760205268859863, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.866969645023346, "num_tokens": 500996746.0, "step": 13136 }, { "epoch": 1.6711614298435313, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.885757446289062, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8619340658187866, "num_tokens": 501034135.0, "step": 13137 }, { "epoch": 1.6712886401221219, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.802531242370605, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8728575110435486, "num_tokens": 501069810.0, "step": 13138 }, { "epoch": 1.6714158504007124, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.828408241271973, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8631985187530518, "num_tokens": 501110597.0, "step": 13139 }, { "epoch": 1.671543060679303, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.800155639648438, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8588477373123169, "num_tokens": 501145927.0, "step": 13140 }, { "epoch": 1.6716702709578934, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.810888290405273, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8502870202064514, "num_tokens": 501188473.0, "step": 13141 }, { "epoch": 1.6717974812364838, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.779722213745117, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8807755708694458, "num_tokens": 501226143.0, "step": 13142 }, { "epoch": 1.6719246915150743, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.820597648620605, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8604104518890381, "num_tokens": 501266980.0, "step": 13143 }, { "epoch": 1.6720519017936648, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.789799690246582, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.865769624710083, "num_tokens": 501305414.0, "step": 13144 }, { "epoch": 1.6721791120722553, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.831575393676758, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8610421419143677, "num_tokens": 501343009.0, "step": 13145 }, { "epoch": 1.6723063223508459, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.781258583068848, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8751339912414551, "num_tokens": 501383571.0, "step": 13146 }, { "epoch": 1.6724335326294364, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.800707817077637, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8703936338424683, "num_tokens": 501425444.0, "step": 13147 }, { "epoch": 1.672560742908027, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.741222381591797, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8816676735877991, "num_tokens": 501466536.0, "step": 13148 }, { "epoch": 1.6726879531866174, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.786735534667969, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8714345693588257, "num_tokens": 501507346.0, "step": 13149 }, { "epoch": 1.672815163465208, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.803587913513184, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.875787079334259, "num_tokens": 501545515.0, "step": 13150 }, { "epoch": 1.6729423737437985, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.707085609436035, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8625843524932861, "num_tokens": 501588793.0, "step": 13151 }, { "epoch": 1.673069584022389, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.800372123718262, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.866004228591919, "num_tokens": 501624384.0, "step": 13152 }, { "epoch": 1.6731967943009796, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.735836029052734, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8682937026023865, "num_tokens": 501663150.0, "step": 13153 }, { "epoch": 1.67332400457957, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.758915901184082, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8619404435157776, "num_tokens": 501706177.0, "step": 13154 }, { "epoch": 1.6734512148581606, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.733830451965332, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8807041645050049, "num_tokens": 501745267.0, "step": 13155 }, { "epoch": 1.6735784251367511, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.801541328430176, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8640928864479065, "num_tokens": 501783591.0, "step": 13156 }, { "epoch": 1.6737056354153417, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.812922477722168, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8594253063201904, "num_tokens": 501825190.0, "step": 13157 }, { "epoch": 1.6738328456939322, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.8001127243042, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8694136738777161, "num_tokens": 501855813.0, "step": 13158 }, { "epoch": 1.6739600559725227, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.831968307495117, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8647931814193726, "num_tokens": 501890106.0, "step": 13159 }, { "epoch": 1.674087266251113, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.749080657958984, "learning_rate": 1e-06, "loss": 0.457, "mean_token_accuracy": 0.855096697807312, "num_tokens": 501935437.0, "step": 13160 }, { "epoch": 1.6742144765297036, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.813262939453125, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8558719158172607, "num_tokens": 501980134.0, "step": 13161 }, { "epoch": 1.674341686808294, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.74806022644043, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8626644611358643, "num_tokens": 502015977.0, "step": 13162 }, { "epoch": 1.6744688970868846, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.7711820602417, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8718858361244202, "num_tokens": 502061717.0, "step": 13163 }, { "epoch": 1.6745961073654752, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.754487991333008, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8665132522583008, "num_tokens": 502101594.0, "step": 13164 }, { "epoch": 1.6747233176440657, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.780609130859375, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8687901496887207, "num_tokens": 502136521.0, "step": 13165 }, { "epoch": 1.674850527922656, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.810168266296387, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8493480682373047, "num_tokens": 502174821.0, "step": 13166 }, { "epoch": 1.6749777382012465, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.737997055053711, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8621935844421387, "num_tokens": 502210615.0, "step": 13167 }, { "epoch": 1.675104948479837, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.80103588104248, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8684276938438416, "num_tokens": 502248220.0, "step": 13168 }, { "epoch": 1.6752321587584276, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.734659194946289, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8666496276855469, "num_tokens": 502290686.0, "step": 13169 }, { "epoch": 1.675359369037018, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.81396770477295, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8564954996109009, "num_tokens": 502331789.0, "step": 13170 }, { "epoch": 1.6754865793156086, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.725298881530762, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8762274980545044, "num_tokens": 502371490.0, "step": 13171 }, { "epoch": 1.6756137895941992, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.747288703918457, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8676354289054871, "num_tokens": 502415606.0, "step": 13172 }, { "epoch": 1.6757409998727897, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.769338607788086, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.867567777633667, "num_tokens": 502456203.0, "step": 13173 }, { "epoch": 1.6758682101513802, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.790582656860352, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8848800659179688, "num_tokens": 502491508.0, "step": 13174 }, { "epoch": 1.6759954204299707, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.753117561340332, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8574297428131104, "num_tokens": 502533408.0, "step": 13175 }, { "epoch": 1.6761226307085613, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.805187225341797, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8614276647567749, "num_tokens": 502569824.0, "step": 13176 }, { "epoch": 1.6762498409871518, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.753392219543457, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8766255378723145, "num_tokens": 502604776.0, "step": 13177 }, { "epoch": 1.6763770512657423, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.773968696594238, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8501235246658325, "num_tokens": 502645732.0, "step": 13178 }, { "epoch": 1.6765042615443329, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.76610279083252, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8716320991516113, "num_tokens": 502686044.0, "step": 13179 }, { "epoch": 1.6766314718229234, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.749346733093262, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8591947555541992, "num_tokens": 502728288.0, "step": 13180 }, { "epoch": 1.676758682101514, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.779115676879883, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8598843812942505, "num_tokens": 502760772.0, "step": 13181 }, { "epoch": 1.6768858923801044, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.796659469604492, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8680351972579956, "num_tokens": 502791585.0, "step": 13182 }, { "epoch": 1.677013102658695, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.757308959960938, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8855020999908447, "num_tokens": 502826339.0, "step": 13183 }, { "epoch": 1.6771403129372855, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.694563865661621, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8853152990341187, "num_tokens": 502868716.0, "step": 13184 }, { "epoch": 1.6772675232158758, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.754523277282715, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8689234256744385, "num_tokens": 502907992.0, "step": 13185 }, { "epoch": 1.6773947334944663, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.784586906433105, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8626768589019775, "num_tokens": 502947197.0, "step": 13186 }, { "epoch": 1.6775219437730569, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.778767585754395, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8768050670623779, "num_tokens": 502985890.0, "step": 13187 }, { "epoch": 1.6776491540516474, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.798982620239258, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8637651205062866, "num_tokens": 503033476.0, "step": 13188 }, { "epoch": 1.677776364330238, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.779902458190918, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8769503235816956, "num_tokens": 503068002.0, "step": 13189 }, { "epoch": 1.6779035746088284, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.779693603515625, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.861572265625, "num_tokens": 503110059.0, "step": 13190 }, { "epoch": 1.6780307848874187, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.743558883666992, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8615916967391968, "num_tokens": 503146617.0, "step": 13191 }, { "epoch": 1.6781579951660093, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.813436508178711, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8660493493080139, "num_tokens": 503186436.0, "step": 13192 }, { "epoch": 1.6782852054445998, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.807101249694824, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8731154203414917, "num_tokens": 503227983.0, "step": 13193 }, { "epoch": 1.6784124157231903, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.691250801086426, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8555476665496826, "num_tokens": 503269301.0, "step": 13194 }, { "epoch": 1.6785396260017809, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.802786827087402, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.850904107093811, "num_tokens": 503308919.0, "step": 13195 }, { "epoch": 1.6786668362803714, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.798750877380371, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8778163194656372, "num_tokens": 503345030.0, "step": 13196 }, { "epoch": 1.678794046558962, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.787761688232422, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8603269457817078, "num_tokens": 503379142.0, "step": 13197 }, { "epoch": 1.6789212568375524, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.764659881591797, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8683972358703613, "num_tokens": 503411810.0, "step": 13198 }, { "epoch": 1.679048467116143, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.746944427490234, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8659986257553101, "num_tokens": 503450266.0, "step": 13199 }, { "epoch": 1.6791756773947335, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.776590347290039, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8759756088256836, "num_tokens": 503491015.0, "step": 13200 }, { "epoch": 1.679302887673324, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.763848304748535, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8760346174240112, "num_tokens": 503531964.0, "step": 13201 }, { "epoch": 1.6794300979519146, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.782181739807129, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8721113204956055, "num_tokens": 503576069.0, "step": 13202 }, { "epoch": 1.679557308230505, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.790369987487793, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8692874312400818, "num_tokens": 503607895.0, "step": 13203 }, { "epoch": 1.6796845185090956, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.858057975769043, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.863269567489624, "num_tokens": 503636891.0, "step": 13204 }, { "epoch": 1.6798117287876861, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.74125862121582, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.882414698600769, "num_tokens": 503682037.0, "step": 13205 }, { "epoch": 1.6799389390662767, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.793744087219238, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8704240322113037, "num_tokens": 503724008.0, "step": 13206 }, { "epoch": 1.6800661493448672, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.795544624328613, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8758153915405273, "num_tokens": 503759008.0, "step": 13207 }, { "epoch": 1.6801933596234577, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.799240112304688, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8648509979248047, "num_tokens": 503796599.0, "step": 13208 }, { "epoch": 1.680320569902048, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.734061241149902, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8661377429962158, "num_tokens": 503835539.0, "step": 13209 }, { "epoch": 1.6804477801806386, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.826504707336426, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8649686574935913, "num_tokens": 503875933.0, "step": 13210 }, { "epoch": 1.680574990459229, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.816438674926758, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8659248352050781, "num_tokens": 503913740.0, "step": 13211 }, { "epoch": 1.6807022007378196, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.812868118286133, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8710362315177917, "num_tokens": 503955105.0, "step": 13212 }, { "epoch": 1.6808294110164101, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.807522773742676, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8678939342498779, "num_tokens": 503987641.0, "step": 13213 }, { "epoch": 1.6809566212950007, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.83039379119873, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8598257303237915, "num_tokens": 504028699.0, "step": 13214 }, { "epoch": 1.681083831573591, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.791715621948242, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8619890213012695, "num_tokens": 504065597.0, "step": 13215 }, { "epoch": 1.6812110418521815, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.759234428405762, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8542724251747131, "num_tokens": 504105968.0, "step": 13216 }, { "epoch": 1.681338252130772, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.783778190612793, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8706678152084351, "num_tokens": 504143743.0, "step": 13217 }, { "epoch": 1.6814654624093626, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.804479598999023, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8704525828361511, "num_tokens": 504177023.0, "step": 13218 }, { "epoch": 1.681592672687953, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.760575294494629, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8662797212600708, "num_tokens": 504210492.0, "step": 13219 }, { "epoch": 1.6817198829665436, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.791894912719727, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8691058158874512, "num_tokens": 504248322.0, "step": 13220 }, { "epoch": 1.6818470932451342, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.831430435180664, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.854344367980957, "num_tokens": 504285915.0, "step": 13221 }, { "epoch": 1.6819743035237247, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.785273551940918, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8664364218711853, "num_tokens": 504325578.0, "step": 13222 }, { "epoch": 1.6821015138023152, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.802324295043945, "learning_rate": 1e-06, "loss": 0.3396, "mean_token_accuracy": 0.888396143913269, "num_tokens": 504362876.0, "step": 13223 }, { "epoch": 1.6822287240809057, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.798103332519531, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8672871589660645, "num_tokens": 504402303.0, "step": 13224 }, { "epoch": 1.6823559343594963, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.818056106567383, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.869527280330658, "num_tokens": 504440039.0, "step": 13225 }, { "epoch": 1.6824831446380868, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.778433799743652, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8597924709320068, "num_tokens": 504485102.0, "step": 13226 }, { "epoch": 1.6826103549166773, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.79676628112793, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8665116429328918, "num_tokens": 504522762.0, "step": 13227 }, { "epoch": 1.6827375651952678, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.789995193481445, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8721502423286438, "num_tokens": 504562818.0, "step": 13228 }, { "epoch": 1.6828647754738584, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.770483016967773, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.858885645866394, "num_tokens": 504602914.0, "step": 13229 }, { "epoch": 1.682991985752449, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.809462547302246, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8751823306083679, "num_tokens": 504641376.0, "step": 13230 }, { "epoch": 1.6831191960310394, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.836265563964844, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8752263188362122, "num_tokens": 504675031.0, "step": 13231 }, { "epoch": 1.68324640630963, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.802436828613281, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8749607801437378, "num_tokens": 504709624.0, "step": 13232 }, { "epoch": 1.6833736165882205, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.798330307006836, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8723232746124268, "num_tokens": 504749310.0, "step": 13233 }, { "epoch": 1.6835008268668108, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.770537376403809, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8689874410629272, "num_tokens": 504791700.0, "step": 13234 }, { "epoch": 1.6836280371454013, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.8488130569458, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8587824106216431, "num_tokens": 504833246.0, "step": 13235 }, { "epoch": 1.6837552474239919, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.784944534301758, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8723930716514587, "num_tokens": 504872901.0, "step": 13236 }, { "epoch": 1.6838824577025824, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.799947738647461, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8627406358718872, "num_tokens": 504906138.0, "step": 13237 }, { "epoch": 1.684009667981173, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.810603141784668, "learning_rate": 1e-06, "loss": 0.4842, "mean_token_accuracy": 0.8487145304679871, "num_tokens": 504947147.0, "step": 13238 }, { "epoch": 1.6841368782597634, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.797548294067383, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8695825934410095, "num_tokens": 504985332.0, "step": 13239 }, { "epoch": 1.6842640885383537, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.848868370056152, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8596968650817871, "num_tokens": 505027995.0, "step": 13240 }, { "epoch": 1.6843912988169443, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.77862548828125, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8709138631820679, "num_tokens": 505065311.0, "step": 13241 }, { "epoch": 1.6845185090955348, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.816122055053711, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8646153807640076, "num_tokens": 505105810.0, "step": 13242 }, { "epoch": 1.6846457193741253, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.809905052185059, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8745834231376648, "num_tokens": 505141003.0, "step": 13243 }, { "epoch": 1.6847729296527159, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.803256034851074, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8576968312263489, "num_tokens": 505188025.0, "step": 13244 }, { "epoch": 1.6849001399313064, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.758816719055176, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8826637268066406, "num_tokens": 505226101.0, "step": 13245 }, { "epoch": 1.685027350209897, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.817360877990723, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8587388396263123, "num_tokens": 505264880.0, "step": 13246 }, { "epoch": 1.6851545604884874, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.828052520751953, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8607295751571655, "num_tokens": 505306208.0, "step": 13247 }, { "epoch": 1.685281770767078, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.787168502807617, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8889682292938232, "num_tokens": 505346024.0, "step": 13248 }, { "epoch": 1.6854089810456685, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.781496047973633, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8903741836547852, "num_tokens": 505383644.0, "step": 13249 }, { "epoch": 1.685536191324259, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.760740280151367, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8721396327018738, "num_tokens": 505420365.0, "step": 13250 }, { "epoch": 1.6856634016028496, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.913247108459473, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.870890736579895, "num_tokens": 505458211.0, "step": 13251 }, { "epoch": 1.68579061188144, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.742149353027344, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8637939095497131, "num_tokens": 505501649.0, "step": 13252 }, { "epoch": 1.6859178221600306, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.851314544677734, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.870519757270813, "num_tokens": 505538763.0, "step": 13253 }, { "epoch": 1.6860450324386211, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.82444953918457, "learning_rate": 1e-06, "loss": 0.345, "mean_token_accuracy": 0.8910728693008423, "num_tokens": 505574562.0, "step": 13254 }, { "epoch": 1.6861722427172117, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.822022438049316, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8823972940444946, "num_tokens": 505609718.0, "step": 13255 }, { "epoch": 1.6862994529958022, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.792083740234375, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8814669847488403, "num_tokens": 505644337.0, "step": 13256 }, { "epoch": 1.6864266632743927, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.853654861450195, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8794795274734497, "num_tokens": 505681929.0, "step": 13257 }, { "epoch": 1.686553873552983, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.762958526611328, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8458212614059448, "num_tokens": 505719373.0, "step": 13258 }, { "epoch": 1.6866810838315736, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 12.824807167053223, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8550224304199219, "num_tokens": 505757767.0, "step": 13259 }, { "epoch": 1.686808294110164, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.798125267028809, "learning_rate": 1e-06, "loss": 0.5384, "mean_token_accuracy": 0.8384442925453186, "num_tokens": 505794777.0, "step": 13260 }, { "epoch": 1.6869355043887546, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.796204566955566, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8758087158203125, "num_tokens": 505832672.0, "step": 13261 }, { "epoch": 1.6870627146673451, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.809473991394043, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.857361912727356, "num_tokens": 505873142.0, "step": 13262 }, { "epoch": 1.6871899249459357, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.790813446044922, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8710337281227112, "num_tokens": 505911994.0, "step": 13263 }, { "epoch": 1.687317135224526, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.796196937561035, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8716318607330322, "num_tokens": 505948874.0, "step": 13264 }, { "epoch": 1.6874443455031165, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.794095993041992, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8748743534088135, "num_tokens": 505989679.0, "step": 13265 }, { "epoch": 1.687571555781707, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.795668601989746, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8771654963493347, "num_tokens": 506033209.0, "step": 13266 }, { "epoch": 1.6876987660602976, "ewc_loss": 0.030517578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 12.841757774353027, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8779482245445251, "num_tokens": 506065379.0, "step": 13267 }, { "epoch": 1.687825976338888, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.853323936462402, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.860666036605835, "num_tokens": 506099484.0, "step": 13268 }, { "epoch": 1.6879531866174786, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.809572219848633, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8524942994117737, "num_tokens": 506136559.0, "step": 13269 }, { "epoch": 1.6880803968960691, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.780035018920898, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8674107789993286, "num_tokens": 506172997.0, "step": 13270 }, { "epoch": 1.6882076071746597, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.784563064575195, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.88327956199646, "num_tokens": 506209858.0, "step": 13271 }, { "epoch": 1.6883348174532502, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.838837623596191, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.880903959274292, "num_tokens": 506242260.0, "step": 13272 }, { "epoch": 1.6884620277318407, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.796126365661621, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8555901050567627, "num_tokens": 506276856.0, "step": 13273 }, { "epoch": 1.6885892380104313, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.7791166305542, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.8871755599975586, "num_tokens": 506311469.0, "step": 13274 }, { "epoch": 1.6887164482890218, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.84461498260498, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8787569999694824, "num_tokens": 506348240.0, "step": 13275 }, { "epoch": 1.6888436585676123, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.748173713684082, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8710740804672241, "num_tokens": 506387203.0, "step": 13276 }, { "epoch": 1.6889708688462028, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.81163501739502, "learning_rate": 1e-06, "loss": 0.4865, "mean_token_accuracy": 0.8465508222579956, "num_tokens": 506421869.0, "step": 13277 }, { "epoch": 1.6890980791247934, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.801838874816895, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8698546290397644, "num_tokens": 506458873.0, "step": 13278 }, { "epoch": 1.689225289403384, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.827674865722656, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8563233017921448, "num_tokens": 506495947.0, "step": 13279 }, { "epoch": 1.6893524996819744, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.799460411071777, "learning_rate": 1e-06, "loss": 0.4672, "mean_token_accuracy": 0.8501712083816528, "num_tokens": 506533098.0, "step": 13280 }, { "epoch": 1.689479709960565, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.777392387390137, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.879249632358551, "num_tokens": 506571346.0, "step": 13281 }, { "epoch": 1.6896069202391555, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.796422958374023, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8706463575363159, "num_tokens": 506607912.0, "step": 13282 }, { "epoch": 1.6897341305177458, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.764420509338379, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8577490448951721, "num_tokens": 506654258.0, "step": 13283 }, { "epoch": 1.6898613407963363, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.793817520141602, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8755075931549072, "num_tokens": 506693436.0, "step": 13284 }, { "epoch": 1.6899885510749268, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.760605812072754, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8712433576583862, "num_tokens": 506737383.0, "step": 13285 }, { "epoch": 1.6901157613535174, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.737391471862793, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8786637783050537, "num_tokens": 506772282.0, "step": 13286 }, { "epoch": 1.690242971632108, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.802948951721191, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8803998231887817, "num_tokens": 506807208.0, "step": 13287 }, { "epoch": 1.6903701819106984, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.771202087402344, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8745633959770203, "num_tokens": 506847952.0, "step": 13288 }, { "epoch": 1.6904973921892887, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.842555046081543, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8728049993515015, "num_tokens": 506888265.0, "step": 13289 }, { "epoch": 1.6906246024678793, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.78199291229248, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8727705478668213, "num_tokens": 506926096.0, "step": 13290 }, { "epoch": 1.6907518127464698, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.765064239501953, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.863268256187439, "num_tokens": 506962252.0, "step": 13291 }, { "epoch": 1.6908790230250603, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.786174774169922, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8749563694000244, "num_tokens": 507004709.0, "step": 13292 }, { "epoch": 1.6910062333036509, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.803872108459473, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8847795724868774, "num_tokens": 507041041.0, "step": 13293 }, { "epoch": 1.6911334435822414, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.830361366271973, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8627889752388, "num_tokens": 507080251.0, "step": 13294 }, { "epoch": 1.691260653860832, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.731164932250977, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8678086996078491, "num_tokens": 507120813.0, "step": 13295 }, { "epoch": 1.6913878641394224, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.788368225097656, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8559496998786926, "num_tokens": 507160284.0, "step": 13296 }, { "epoch": 1.691515074418013, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.770100593566895, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8624265193939209, "num_tokens": 507205069.0, "step": 13297 }, { "epoch": 1.6916422846966035, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.77450180053711, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8775366544723511, "num_tokens": 507244689.0, "step": 13298 }, { "epoch": 1.691769494975194, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.788721084594727, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8710268139839172, "num_tokens": 507283917.0, "step": 13299 }, { "epoch": 1.6918967052537845, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.74726390838623, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8715416789054871, "num_tokens": 507327937.0, "step": 13300 }, { "epoch": 1.692023915532375, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.739151954650879, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8834933042526245, "num_tokens": 507369521.0, "step": 13301 }, { "epoch": 1.6921511258109656, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.783686637878418, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8797135353088379, "num_tokens": 507408157.0, "step": 13302 }, { "epoch": 1.6922783360895561, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.776928901672363, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8751199245452881, "num_tokens": 507448003.0, "step": 13303 }, { "epoch": 1.6924055463681467, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.705290794372559, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8768537044525146, "num_tokens": 507489970.0, "step": 13304 }, { "epoch": 1.6925327566467372, "ewc_loss": 0.03076171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.779901504516602, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8706328868865967, "num_tokens": 507524536.0, "step": 13305 }, { "epoch": 1.6926599669253277, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.799127578735352, "learning_rate": 1e-06, "loss": 0.3334, "mean_token_accuracy": 0.8890830278396606, "num_tokens": 507561575.0, "step": 13306 }, { "epoch": 1.692787177203918, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.745368003845215, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.872029721736908, "num_tokens": 507597212.0, "step": 13307 }, { "epoch": 1.6929143874825086, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.772012710571289, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8719576001167297, "num_tokens": 507634251.0, "step": 13308 }, { "epoch": 1.693041597761099, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.776896476745605, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8755974173545837, "num_tokens": 507671345.0, "step": 13309 }, { "epoch": 1.6931688080396896, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.785246849060059, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8591659665107727, "num_tokens": 507709797.0, "step": 13310 }, { "epoch": 1.6932960183182801, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.761080741882324, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8691387176513672, "num_tokens": 507747047.0, "step": 13311 }, { "epoch": 1.6934232285968707, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.73851203918457, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8700053691864014, "num_tokens": 507782187.0, "step": 13312 }, { "epoch": 1.693550438875461, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.765119552612305, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8567943572998047, "num_tokens": 507826714.0, "step": 13313 }, { "epoch": 1.6936776491540515, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.808732032775879, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8615410327911377, "num_tokens": 507869423.0, "step": 13314 }, { "epoch": 1.693804859432642, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.834504127502441, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8792033195495605, "num_tokens": 507904711.0, "step": 13315 }, { "epoch": 1.6939320697112326, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.842447280883789, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8640723824501038, "num_tokens": 507947760.0, "step": 13316 }, { "epoch": 1.694059279989823, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.78309154510498, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8665876388549805, "num_tokens": 507984249.0, "step": 13317 }, { "epoch": 1.6941864902684136, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.891862869262695, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8662039041519165, "num_tokens": 508017206.0, "step": 13318 }, { "epoch": 1.6943137005470041, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.814279556274414, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8639672994613647, "num_tokens": 508056357.0, "step": 13319 }, { "epoch": 1.6944409108255947, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.855567932128906, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8610131740570068, "num_tokens": 508094870.0, "step": 13320 }, { "epoch": 1.6945681211041852, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.853221893310547, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8684030175209045, "num_tokens": 508120323.0, "step": 13321 }, { "epoch": 1.6946953313827757, "ewc_loss": 0.031005859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.816221237182617, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8608657717704773, "num_tokens": 508160018.0, "step": 13322 }, { "epoch": 1.6948225416613663, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.850435256958008, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8867555260658264, "num_tokens": 508200947.0, "step": 13323 }, { "epoch": 1.6949497519399568, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.810521125793457, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.856147289276123, "num_tokens": 508246457.0, "step": 13324 }, { "epoch": 1.6950769622185473, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.865180015563965, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8641614317893982, "num_tokens": 508281951.0, "step": 13325 }, { "epoch": 1.6952041724971378, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.818453788757324, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8503800630569458, "num_tokens": 508322571.0, "step": 13326 }, { "epoch": 1.6953313827757284, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.85676383972168, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8767573237419128, "num_tokens": 508357347.0, "step": 13327 }, { "epoch": 1.695458593054319, "ewc_loss": 0.03125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.758079528808594, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.871263325214386, "num_tokens": 508401813.0, "step": 13328 }, { "epoch": 1.6955858033329094, "ewc_loss": 0.031494140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.894062042236328, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.857642412185669, "num_tokens": 508440457.0, "step": 13329 }, { "epoch": 1.6957130136115, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.879413604736328, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8598988652229309, "num_tokens": 508472272.0, "step": 13330 }, { "epoch": 1.6958402238900905, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.8815336227417, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8816086053848267, "num_tokens": 508509352.0, "step": 13331 }, { "epoch": 1.6959674341686808, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.853378295898438, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8781286478042603, "num_tokens": 508552515.0, "step": 13332 }, { "epoch": 1.6960946444472713, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.862382888793945, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8815907835960388, "num_tokens": 508591399.0, "step": 13333 }, { "epoch": 1.6962218547258618, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.78357219696045, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8741312623023987, "num_tokens": 508630427.0, "step": 13334 }, { "epoch": 1.6963490650044524, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.91409683227539, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8768236041069031, "num_tokens": 508667469.0, "step": 13335 }, { "epoch": 1.696476275283043, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.818441390991211, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8724390268325806, "num_tokens": 508706308.0, "step": 13336 }, { "epoch": 1.6966034855616334, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.881193161010742, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.857670783996582, "num_tokens": 508738968.0, "step": 13337 }, { "epoch": 1.6967306958402237, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.844155311584473, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.875487208366394, "num_tokens": 508769879.0, "step": 13338 }, { "epoch": 1.6968579061188143, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.80910587310791, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8534711599349976, "num_tokens": 508812277.0, "step": 13339 }, { "epoch": 1.6969851163974048, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.86963176727295, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8591440320014954, "num_tokens": 508849293.0, "step": 13340 }, { "epoch": 1.6971123266759953, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.846282958984375, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8709776997566223, "num_tokens": 508885798.0, "step": 13341 }, { "epoch": 1.6972395369545858, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.864022254943848, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8673124313354492, "num_tokens": 508920491.0, "step": 13342 }, { "epoch": 1.6973667472331764, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.877705574035645, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8683687448501587, "num_tokens": 508956161.0, "step": 13343 }, { "epoch": 1.697493957511767, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.815225601196289, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8661766648292542, "num_tokens": 508995696.0, "step": 13344 }, { "epoch": 1.6976211677903574, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.837625503540039, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8746219873428345, "num_tokens": 509029505.0, "step": 13345 }, { "epoch": 1.697748378068948, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.8198823928833, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8775699734687805, "num_tokens": 509062815.0, "step": 13346 }, { "epoch": 1.6978755883475385, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.854848861694336, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8754785656929016, "num_tokens": 509103376.0, "step": 13347 }, { "epoch": 1.698002798626129, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.796395301818848, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8800140619277954, "num_tokens": 509142797.0, "step": 13348 }, { "epoch": 1.6981300089047195, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.847479820251465, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8580880761146545, "num_tokens": 509182724.0, "step": 13349 }, { "epoch": 1.69825721918331, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.813018798828125, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8778959512710571, "num_tokens": 509217856.0, "step": 13350 }, { "epoch": 1.6983844294619006, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.88081169128418, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8702499866485596, "num_tokens": 509251581.0, "step": 13351 }, { "epoch": 1.6985116397404911, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.796987533569336, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8857166767120361, "num_tokens": 509286687.0, "step": 13352 }, { "epoch": 1.6986388500190817, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.773050308227539, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.864096999168396, "num_tokens": 509323372.0, "step": 13353 }, { "epoch": 1.6987660602976722, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.811746597290039, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8615279197692871, "num_tokens": 509362472.0, "step": 13354 }, { "epoch": 1.6988932705762627, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.834613800048828, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8706075549125671, "num_tokens": 509397676.0, "step": 13355 }, { "epoch": 1.699020480854853, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.7482271194458, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8726162910461426, "num_tokens": 509441100.0, "step": 13356 }, { "epoch": 1.6991476911334435, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.824191093444824, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8651930689811707, "num_tokens": 509485430.0, "step": 13357 }, { "epoch": 1.699274901412034, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.871478080749512, "learning_rate": 1e-06, "loss": 0.5074, "mean_token_accuracy": 0.840302050113678, "num_tokens": 509520102.0, "step": 13358 }, { "epoch": 1.6994021116906246, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.852277755737305, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.871015727519989, "num_tokens": 509556525.0, "step": 13359 }, { "epoch": 1.6995293219692151, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.83710765838623, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8630193471908569, "num_tokens": 509596899.0, "step": 13360 }, { "epoch": 1.6996565322478057, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.842290878295898, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8697307109832764, "num_tokens": 509632125.0, "step": 13361 }, { "epoch": 1.699783742526396, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.809951782226562, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8652018904685974, "num_tokens": 509668131.0, "step": 13362 }, { "epoch": 1.6999109528049865, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.861928939819336, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8817344903945923, "num_tokens": 509704078.0, "step": 13363 }, { "epoch": 1.700038163083577, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.803670883178711, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8823709487915039, "num_tokens": 509739464.0, "step": 13364 }, { "epoch": 1.7001653733621676, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.866159439086914, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8680172562599182, "num_tokens": 509777970.0, "step": 13365 }, { "epoch": 1.700292583640758, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.868024826049805, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8548682928085327, "num_tokens": 509817010.0, "step": 13366 }, { "epoch": 1.7004197939193486, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.882762908935547, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8747381567955017, "num_tokens": 509854192.0, "step": 13367 }, { "epoch": 1.7005470041979391, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.849223136901855, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8834218978881836, "num_tokens": 509888633.0, "step": 13368 }, { "epoch": 1.7006742144765297, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.855229377746582, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.870439887046814, "num_tokens": 509928474.0, "step": 13369 }, { "epoch": 1.7008014247551202, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.831850051879883, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.862781822681427, "num_tokens": 509968875.0, "step": 13370 }, { "epoch": 1.7009286350337107, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.825182914733887, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8700808882713318, "num_tokens": 510006021.0, "step": 13371 }, { "epoch": 1.7010558453123013, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.798612594604492, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8749907612800598, "num_tokens": 510042111.0, "step": 13372 }, { "epoch": 1.7011830555908918, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.858884811401367, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8748458623886108, "num_tokens": 510074894.0, "step": 13373 }, { "epoch": 1.7013102658694823, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.831750869750977, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.861343502998352, "num_tokens": 510111516.0, "step": 13374 }, { "epoch": 1.7014374761480728, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.815072059631348, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8717986345291138, "num_tokens": 510151549.0, "step": 13375 }, { "epoch": 1.7015646864266634, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.823897361755371, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8694599866867065, "num_tokens": 510189215.0, "step": 13376 }, { "epoch": 1.701691896705254, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.776880264282227, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.8884352445602417, "num_tokens": 510231189.0, "step": 13377 }, { "epoch": 1.7018191069838444, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.827842712402344, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8733636140823364, "num_tokens": 510273956.0, "step": 13378 }, { "epoch": 1.701946317262435, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.796041488647461, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8853394389152527, "num_tokens": 510315095.0, "step": 13379 }, { "epoch": 1.7020735275410255, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.852846145629883, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8502174615859985, "num_tokens": 510350753.0, "step": 13380 }, { "epoch": 1.7022007378196158, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.85197925567627, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8741685152053833, "num_tokens": 510389257.0, "step": 13381 }, { "epoch": 1.7023279480982063, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.904949188232422, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8700819611549377, "num_tokens": 510428879.0, "step": 13382 }, { "epoch": 1.7024551583767968, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.806551933288574, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8727642297744751, "num_tokens": 510468409.0, "step": 13383 }, { "epoch": 1.7025823686553874, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.82152271270752, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8846818208694458, "num_tokens": 510503988.0, "step": 13384 }, { "epoch": 1.702709578933978, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.895219802856445, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8675190210342407, "num_tokens": 510539009.0, "step": 13385 }, { "epoch": 1.7028367892125684, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.859452247619629, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8844772577285767, "num_tokens": 510576177.0, "step": 13386 }, { "epoch": 1.7029639994911587, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.82627010345459, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8693943023681641, "num_tokens": 510622194.0, "step": 13387 }, { "epoch": 1.7030912097697493, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.840027809143066, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8866966962814331, "num_tokens": 510663564.0, "step": 13388 }, { "epoch": 1.7032184200483398, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.858404159545898, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8806869983673096, "num_tokens": 510702153.0, "step": 13389 }, { "epoch": 1.7033456303269303, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.818745613098145, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8661574125289917, "num_tokens": 510742812.0, "step": 13390 }, { "epoch": 1.7034728406055208, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.874650001525879, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8647661805152893, "num_tokens": 510783363.0, "step": 13391 }, { "epoch": 1.7036000508841114, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.828390121459961, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8558596968650818, "num_tokens": 510818362.0, "step": 13392 }, { "epoch": 1.703727261162702, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.84998893737793, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8822451829910278, "num_tokens": 510854732.0, "step": 13393 }, { "epoch": 1.7038544714412924, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.846199035644531, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8637030124664307, "num_tokens": 510895107.0, "step": 13394 }, { "epoch": 1.703981681719883, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.845998764038086, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8672255873680115, "num_tokens": 510930737.0, "step": 13395 }, { "epoch": 1.7041088919984735, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.895673751831055, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8778026103973389, "num_tokens": 510964646.0, "step": 13396 }, { "epoch": 1.704236102277064, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.84586238861084, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8733100891113281, "num_tokens": 511000583.0, "step": 13397 }, { "epoch": 1.7043633125556545, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.877553939819336, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8721120357513428, "num_tokens": 511033729.0, "step": 13398 }, { "epoch": 1.704490522834245, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.858529090881348, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8752302527427673, "num_tokens": 511074099.0, "step": 13399 }, { "epoch": 1.7046177331128356, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.864664077758789, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8573516011238098, "num_tokens": 511112473.0, "step": 13400 }, { "epoch": 1.7047449433914261, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.88254451751709, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8705151081085205, "num_tokens": 511149839.0, "step": 13401 }, { "epoch": 1.7048721536700167, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.87033462524414, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8592094779014587, "num_tokens": 511183054.0, "step": 13402 }, { "epoch": 1.7049993639486072, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.900369644165039, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8741382956504822, "num_tokens": 511218498.0, "step": 13403 }, { "epoch": 1.7051265742271977, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.82366943359375, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8666979074478149, "num_tokens": 511259776.0, "step": 13404 }, { "epoch": 1.705253784505788, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.92778205871582, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8767431974411011, "num_tokens": 511297331.0, "step": 13405 }, { "epoch": 1.7053809947843785, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.792768478393555, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.882124662399292, "num_tokens": 511331055.0, "step": 13406 }, { "epoch": 1.705508205062969, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.83167839050293, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8885552883148193, "num_tokens": 511368651.0, "step": 13407 }, { "epoch": 1.7056354153415596, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.845626831054688, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8747445344924927, "num_tokens": 511402163.0, "step": 13408 }, { "epoch": 1.7057626256201501, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.858416557312012, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8640639781951904, "num_tokens": 511443769.0, "step": 13409 }, { "epoch": 1.7058898358987407, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.802757263183594, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8671534061431885, "num_tokens": 511481038.0, "step": 13410 }, { "epoch": 1.706017046177331, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.83219051361084, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8713148832321167, "num_tokens": 511522940.0, "step": 13411 }, { "epoch": 1.7061442564559215, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.800714492797852, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8671325445175171, "num_tokens": 511562630.0, "step": 13412 }, { "epoch": 1.706271466734512, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.90098762512207, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8558374643325806, "num_tokens": 511598986.0, "step": 13413 }, { "epoch": 1.7063986770131025, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.848594665527344, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8719379901885986, "num_tokens": 511633582.0, "step": 13414 }, { "epoch": 1.706525887291693, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.808148384094238, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.870380163192749, "num_tokens": 511673296.0, "step": 13415 }, { "epoch": 1.7066530975702836, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.976187705993652, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8619481325149536, "num_tokens": 511707890.0, "step": 13416 }, { "epoch": 1.7067803078488741, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.844822883605957, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.875884473323822, "num_tokens": 511741597.0, "step": 13417 }, { "epoch": 1.7069075181274647, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.879570960998535, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8737716674804688, "num_tokens": 511776816.0, "step": 13418 }, { "epoch": 1.7070347284060552, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.855306625366211, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8635584115982056, "num_tokens": 511814866.0, "step": 13419 }, { "epoch": 1.7071619386846457, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.813231468200684, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8714039325714111, "num_tokens": 511852458.0, "step": 13420 }, { "epoch": 1.7072891489632362, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.87135124206543, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8647526502609253, "num_tokens": 511891201.0, "step": 13421 }, { "epoch": 1.7074163592418268, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.834099769592285, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.869813859462738, "num_tokens": 511930862.0, "step": 13422 }, { "epoch": 1.7075435695204173, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.820721626281738, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8773615956306458, "num_tokens": 511967695.0, "step": 13423 }, { "epoch": 1.7076707797990078, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.819226264953613, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8838034868240356, "num_tokens": 512005976.0, "step": 13424 }, { "epoch": 1.7077979900775984, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.850204467773438, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8716258406639099, "num_tokens": 512041575.0, "step": 13425 }, { "epoch": 1.7079252003561889, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.825450897216797, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8662219047546387, "num_tokens": 512076171.0, "step": 13426 }, { "epoch": 1.7080524106347794, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.85916519165039, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8664159774780273, "num_tokens": 512112082.0, "step": 13427 }, { "epoch": 1.70817962091337, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.85319995880127, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8563008308410645, "num_tokens": 512149126.0, "step": 13428 }, { "epoch": 1.7083068311919605, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.863219261169434, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8716840744018555, "num_tokens": 512182368.0, "step": 13429 }, { "epoch": 1.7084340414705508, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.890031814575195, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8508447408676147, "num_tokens": 512218224.0, "step": 13430 }, { "epoch": 1.7085612517491413, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.881685256958008, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8689842224121094, "num_tokens": 512255523.0, "step": 13431 }, { "epoch": 1.7086884620277318, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.925292015075684, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8797481060028076, "num_tokens": 512290514.0, "step": 13432 }, { "epoch": 1.7088156723063224, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.823385238647461, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.879101574420929, "num_tokens": 512328468.0, "step": 13433 }, { "epoch": 1.708942882584913, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.861038208007812, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8703766465187073, "num_tokens": 512359786.0, "step": 13434 }, { "epoch": 1.7090700928635034, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.927412986755371, "learning_rate": 1e-06, "loss": 0.4891, "mean_token_accuracy": 0.8488975763320923, "num_tokens": 512395679.0, "step": 13435 }, { "epoch": 1.7091973031420937, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.812699317932129, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8675453066825867, "num_tokens": 512433491.0, "step": 13436 }, { "epoch": 1.7093245134206843, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.910564422607422, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8644438982009888, "num_tokens": 512470215.0, "step": 13437 }, { "epoch": 1.7094517236992748, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.85326099395752, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8652370572090149, "num_tokens": 512507055.0, "step": 13438 }, { "epoch": 1.7095789339778653, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.912721633911133, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8521815538406372, "num_tokens": 512547877.0, "step": 13439 }, { "epoch": 1.7097061442564558, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.86054801940918, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8567375540733337, "num_tokens": 512586558.0, "step": 13440 }, { "epoch": 1.7098333545350464, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.847956657409668, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8671650886535645, "num_tokens": 512631181.0, "step": 13441 }, { "epoch": 1.709960564813637, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.891613006591797, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8801284432411194, "num_tokens": 512667925.0, "step": 13442 }, { "epoch": 1.7100877750922274, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.937223434448242, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8521806001663208, "num_tokens": 512703136.0, "step": 13443 }, { "epoch": 1.710214985370818, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.851720809936523, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8727877140045166, "num_tokens": 512739001.0, "step": 13444 }, { "epoch": 1.7103421956494085, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.823225021362305, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8734421730041504, "num_tokens": 512780517.0, "step": 13445 }, { "epoch": 1.710469405927999, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.866780281066895, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8760122060775757, "num_tokens": 512815964.0, "step": 13446 }, { "epoch": 1.7105966162065895, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.898637771606445, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8724119663238525, "num_tokens": 512855451.0, "step": 13447 }, { "epoch": 1.71072382648518, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.898037910461426, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8734396696090698, "num_tokens": 512895809.0, "step": 13448 }, { "epoch": 1.7108510367637706, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.85578727722168, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8740918636322021, "num_tokens": 512936424.0, "step": 13449 }, { "epoch": 1.7109782470423611, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.886232376098633, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8844900727272034, "num_tokens": 512976486.0, "step": 13450 }, { "epoch": 1.7111054573209517, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.893375396728516, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8613466024398804, "num_tokens": 513014131.0, "step": 13451 }, { "epoch": 1.7112326675995422, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.86320686340332, "learning_rate": 1e-06, "loss": 0.4928, "mean_token_accuracy": 0.842280387878418, "num_tokens": 513054706.0, "step": 13452 }, { "epoch": 1.7113598778781327, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.8900785446167, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8745903372764587, "num_tokens": 513094089.0, "step": 13453 }, { "epoch": 1.711487088156723, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.834188461303711, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8677594661712646, "num_tokens": 513136306.0, "step": 13454 }, { "epoch": 1.7116142984353135, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.893403053283691, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8649656772613525, "num_tokens": 513174376.0, "step": 13455 }, { "epoch": 1.711741508713904, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.839947700500488, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8731048107147217, "num_tokens": 513211834.0, "step": 13456 }, { "epoch": 1.7118687189924946, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.906542778015137, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.861573338508606, "num_tokens": 513241765.0, "step": 13457 }, { "epoch": 1.7119959292710851, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.905777931213379, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8565291166305542, "num_tokens": 513283510.0, "step": 13458 }, { "epoch": 1.7121231395496757, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.936394691467285, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8659642934799194, "num_tokens": 513320746.0, "step": 13459 }, { "epoch": 1.712250349828266, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.892651557922363, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8668137192726135, "num_tokens": 513358036.0, "step": 13460 }, { "epoch": 1.7123775601068565, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.821887969970703, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8747172355651855, "num_tokens": 513395156.0, "step": 13461 }, { "epoch": 1.712504770385447, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.860151290893555, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8633967638015747, "num_tokens": 513438631.0, "step": 13462 }, { "epoch": 1.7126319806640375, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.838595390319824, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8606067895889282, "num_tokens": 513471225.0, "step": 13463 }, { "epoch": 1.712759190942628, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.865128517150879, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.861016571521759, "num_tokens": 513513824.0, "step": 13464 }, { "epoch": 1.7128864012212186, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.810303688049316, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8631947636604309, "num_tokens": 513556263.0, "step": 13465 }, { "epoch": 1.7130136114998091, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.875396728515625, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8736977577209473, "num_tokens": 513595093.0, "step": 13466 }, { "epoch": 1.7131408217783997, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.851175308227539, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.861469030380249, "num_tokens": 513633761.0, "step": 13467 }, { "epoch": 1.7132680320569902, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.852909088134766, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8652313947677612, "num_tokens": 513671404.0, "step": 13468 }, { "epoch": 1.7133952423355807, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.867057800292969, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.868694543838501, "num_tokens": 513713752.0, "step": 13469 }, { "epoch": 1.7135224526141712, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.877005577087402, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8654735684394836, "num_tokens": 513755261.0, "step": 13470 }, { "epoch": 1.7136496628927618, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.820780754089355, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8738294243812561, "num_tokens": 513799042.0, "step": 13471 }, { "epoch": 1.7137768731713523, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.803071975708008, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.858769416809082, "num_tokens": 513838148.0, "step": 13472 }, { "epoch": 1.7139040834499428, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.881808280944824, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8757106065750122, "num_tokens": 513878960.0, "step": 13473 }, { "epoch": 1.7140312937285334, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.823675155639648, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8701702356338501, "num_tokens": 513914083.0, "step": 13474 }, { "epoch": 1.7141585040071239, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.905649185180664, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8583028316497803, "num_tokens": 513951591.0, "step": 13475 }, { "epoch": 1.7142857142857144, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.881555557250977, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8564559817314148, "num_tokens": 513986127.0, "step": 13476 }, { "epoch": 1.714412924564305, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.817046165466309, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8639775514602661, "num_tokens": 514028248.0, "step": 13477 }, { "epoch": 1.7145401348428955, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.88627815246582, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.873087465763092, "num_tokens": 514062757.0, "step": 13478 }, { "epoch": 1.7146673451214858, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.852478981018066, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8559330701828003, "num_tokens": 514102635.0, "step": 13479 }, { "epoch": 1.7147945554000763, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.888433456420898, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8678024411201477, "num_tokens": 514140883.0, "step": 13480 }, { "epoch": 1.7149217656786668, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.815882682800293, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8776024580001831, "num_tokens": 514176536.0, "step": 13481 }, { "epoch": 1.7150489759572574, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.897894859313965, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8640775084495544, "num_tokens": 514209969.0, "step": 13482 }, { "epoch": 1.7151761862358479, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.843900680541992, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.868475615978241, "num_tokens": 514249679.0, "step": 13483 }, { "epoch": 1.7153033965144384, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.865396499633789, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8626837730407715, "num_tokens": 514292719.0, "step": 13484 }, { "epoch": 1.7154306067930287, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.916308403015137, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8637787699699402, "num_tokens": 514323396.0, "step": 13485 }, { "epoch": 1.7155578170716193, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.819461822509766, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8739382028579712, "num_tokens": 514360951.0, "step": 13486 }, { "epoch": 1.7156850273502098, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.915144920349121, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8533705472946167, "num_tokens": 514403879.0, "step": 13487 }, { "epoch": 1.7158122376288003, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.858683586120605, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8761916756629944, "num_tokens": 514446628.0, "step": 13488 }, { "epoch": 1.7159394479073908, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.868718147277832, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8700936436653137, "num_tokens": 514486132.0, "step": 13489 }, { "epoch": 1.7160666581859814, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.841994285583496, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8578073382377625, "num_tokens": 514526384.0, "step": 13490 }, { "epoch": 1.716193868464572, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.935826301574707, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8671883344650269, "num_tokens": 514565828.0, "step": 13491 }, { "epoch": 1.7163210787431624, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.886612892150879, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8638797998428345, "num_tokens": 514605260.0, "step": 13492 }, { "epoch": 1.716448289021753, "ewc_loss": 0.03076171875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 12.848601341247559, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.870561420917511, "num_tokens": 514643433.0, "step": 13493 }, { "epoch": 1.7165754993003435, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.837535858154297, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8710619807243347, "num_tokens": 514683264.0, "step": 13494 }, { "epoch": 1.716702709578934, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.87669849395752, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.875306248664856, "num_tokens": 514719775.0, "step": 13495 }, { "epoch": 1.7168299198575245, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.770983695983887, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8616728186607361, "num_tokens": 514766287.0, "step": 13496 }, { "epoch": 1.716957130136115, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.841130256652832, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8776767253875732, "num_tokens": 514806725.0, "step": 13497 }, { "epoch": 1.7170843404147056, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.852420806884766, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8717648386955261, "num_tokens": 514844008.0, "step": 13498 }, { "epoch": 1.7172115506932961, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.841219902038574, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8573707938194275, "num_tokens": 514880356.0, "step": 13499 }, { "epoch": 1.7173387609718866, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 12.857890129089355, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.879847526550293, "num_tokens": 514915309.0, "step": 13500 }, { "epoch": 1.7174659712504772, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.787095069885254, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8619945049285889, "num_tokens": 514956496.0, "step": 13501 }, { "epoch": 1.7175931815290677, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.837762832641602, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8691576719284058, "num_tokens": 514992247.0, "step": 13502 }, { "epoch": 1.717720391807658, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.838093757629395, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8604013919830322, "num_tokens": 515035443.0, "step": 13503 }, { "epoch": 1.7178476020862485, "ewc_loss": 0.031005859375, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.820137023925781, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8728458881378174, "num_tokens": 515072279.0, "step": 13504 }, { "epoch": 1.717974812364839, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.85455322265625, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8679701089859009, "num_tokens": 515112547.0, "step": 13505 }, { "epoch": 1.7181020226434296, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.84044361114502, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8568962216377258, "num_tokens": 515146814.0, "step": 13506 }, { "epoch": 1.7182292329220201, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.855656623840332, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8704873323440552, "num_tokens": 515184393.0, "step": 13507 }, { "epoch": 1.7183564432006107, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.816237449645996, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.861264705657959, "num_tokens": 515225312.0, "step": 13508 }, { "epoch": 1.718483653479201, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.859665870666504, "learning_rate": 1e-06, "loss": 0.484, "mean_token_accuracy": 0.8464484214782715, "num_tokens": 515261330.0, "step": 13509 }, { "epoch": 1.7186108637577915, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.855920791625977, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8745036125183105, "num_tokens": 515299325.0, "step": 13510 }, { "epoch": 1.718738074036382, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.846956253051758, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8689119815826416, "num_tokens": 515335213.0, "step": 13511 }, { "epoch": 1.7188652843149725, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.877840995788574, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8797239065170288, "num_tokens": 515365699.0, "step": 13512 }, { "epoch": 1.718992494593563, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.839089393615723, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8780364990234375, "num_tokens": 515396025.0, "step": 13513 }, { "epoch": 1.7191197048721536, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.856999397277832, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8689303994178772, "num_tokens": 515429724.0, "step": 13514 }, { "epoch": 1.7192469151507441, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.82921028137207, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8650603294372559, "num_tokens": 515465894.0, "step": 13515 }, { "epoch": 1.7193741254293347, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.878108978271484, "learning_rate": 1e-06, "loss": 0.3536, "mean_token_accuracy": 0.887254536151886, "num_tokens": 515498894.0, "step": 13516 }, { "epoch": 1.7195013357079252, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.85078239440918, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8683174252510071, "num_tokens": 515538783.0, "step": 13517 }, { "epoch": 1.7196285459865157, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.884016036987305, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8673543930053711, "num_tokens": 515573107.0, "step": 13518 }, { "epoch": 1.7197557562651062, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.852910041809082, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8703936338424683, "num_tokens": 515614901.0, "step": 13519 }, { "epoch": 1.7198829665436968, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.893402099609375, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8712356090545654, "num_tokens": 515654197.0, "step": 13520 }, { "epoch": 1.7200101768222873, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.82624340057373, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8877773880958557, "num_tokens": 515697003.0, "step": 13521 }, { "epoch": 1.7201373871008778, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.862292289733887, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.865776002407074, "num_tokens": 515737801.0, "step": 13522 }, { "epoch": 1.7202645973794684, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.925628662109375, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8611918687820435, "num_tokens": 515774711.0, "step": 13523 }, { "epoch": 1.7203918076580589, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.851049423217773, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8605132699012756, "num_tokens": 515811728.0, "step": 13524 }, { "epoch": 1.7205190179366494, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.859658241271973, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8739820718765259, "num_tokens": 515844425.0, "step": 13525 }, { "epoch": 1.72064622821524, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.89810562133789, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8711637854576111, "num_tokens": 515882439.0, "step": 13526 }, { "epoch": 1.7207734384938305, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.879044532775879, "learning_rate": 1e-06, "loss": 0.4883, "mean_token_accuracy": 0.8468510508537292, "num_tokens": 515922168.0, "step": 13527 }, { "epoch": 1.7209006487724208, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.88853931427002, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8756915926933289, "num_tokens": 515961840.0, "step": 13528 }, { "epoch": 1.7210278590510113, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.82949447631836, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.86858069896698, "num_tokens": 515997586.0, "step": 13529 }, { "epoch": 1.7211550693296018, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.842906951904297, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8746931552886963, "num_tokens": 516036660.0, "step": 13530 }, { "epoch": 1.7212822796081924, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.8740816116333, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.862017035484314, "num_tokens": 516070846.0, "step": 13531 }, { "epoch": 1.7214094898867829, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.846664428710938, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8904693722724915, "num_tokens": 516107027.0, "step": 13532 }, { "epoch": 1.7215367001653734, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.845956802368164, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8695778250694275, "num_tokens": 516143971.0, "step": 13533 }, { "epoch": 1.7216639104439637, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.870938301086426, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8713895678520203, "num_tokens": 516180617.0, "step": 13534 }, { "epoch": 1.7217911207225542, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.843033790588379, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8698180913925171, "num_tokens": 516220371.0, "step": 13535 }, { "epoch": 1.7219183310011448, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.89542007446289, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8714213371276855, "num_tokens": 516258642.0, "step": 13536 }, { "epoch": 1.7220455412797353, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.810471534729004, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8628833889961243, "num_tokens": 516298971.0, "step": 13537 }, { "epoch": 1.7221727515583258, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.888480186462402, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8808884620666504, "num_tokens": 516334457.0, "step": 13538 }, { "epoch": 1.7222999618369164, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.819051742553711, "learning_rate": 1e-06, "loss": 0.3474, "mean_token_accuracy": 0.8939220905303955, "num_tokens": 516374585.0, "step": 13539 }, { "epoch": 1.7224271721155069, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.836409568786621, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.854888916015625, "num_tokens": 516414180.0, "step": 13540 }, { "epoch": 1.7225543823940974, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.865813255310059, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8773650527000427, "num_tokens": 516450189.0, "step": 13541 }, { "epoch": 1.722681592672688, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.874526023864746, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8666867017745972, "num_tokens": 516487549.0, "step": 13542 }, { "epoch": 1.7228088029512785, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.855758666992188, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8561896085739136, "num_tokens": 516526649.0, "step": 13543 }, { "epoch": 1.722936013229869, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.88626480102539, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8744038343429565, "num_tokens": 516563829.0, "step": 13544 }, { "epoch": 1.7230632235084595, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.855562210083008, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8529902100563049, "num_tokens": 516605179.0, "step": 13545 }, { "epoch": 1.72319043378705, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.870868682861328, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8671979904174805, "num_tokens": 516641576.0, "step": 13546 }, { "epoch": 1.7233176440656406, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.907431602478027, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.866855800151825, "num_tokens": 516678687.0, "step": 13547 }, { "epoch": 1.7234448543442311, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.854798316955566, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.86374831199646, "num_tokens": 516711484.0, "step": 13548 }, { "epoch": 1.7235720646228216, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.859664916992188, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8653692007064819, "num_tokens": 516750598.0, "step": 13549 }, { "epoch": 1.7236992749014122, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.863075256347656, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8761795163154602, "num_tokens": 516793438.0, "step": 13550 }, { "epoch": 1.7238264851800027, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.879790306091309, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8834017515182495, "num_tokens": 516828692.0, "step": 13551 }, { "epoch": 1.723953695458593, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.879515647888184, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8701047897338867, "num_tokens": 516867592.0, "step": 13552 }, { "epoch": 1.7240809057371835, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.83182144165039, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8569347262382507, "num_tokens": 516907668.0, "step": 13553 }, { "epoch": 1.724208116015774, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.892730712890625, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8665245771408081, "num_tokens": 516944628.0, "step": 13554 }, { "epoch": 1.7243353262943646, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.859074592590332, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8642895817756653, "num_tokens": 516989027.0, "step": 13555 }, { "epoch": 1.7244625365729551, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.90246295928955, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8641740083694458, "num_tokens": 517024464.0, "step": 13556 }, { "epoch": 1.7245897468515456, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.850234985351562, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8725273013114929, "num_tokens": 517067714.0, "step": 13557 }, { "epoch": 1.724716957130136, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.824832916259766, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8724308609962463, "num_tokens": 517107564.0, "step": 13558 }, { "epoch": 1.7248441674087265, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.832161903381348, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8752123117446899, "num_tokens": 517148240.0, "step": 13559 }, { "epoch": 1.724971377687317, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.855749130249023, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8615800738334656, "num_tokens": 517187841.0, "step": 13560 }, { "epoch": 1.7250985879659075, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.85966968536377, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.852284848690033, "num_tokens": 517225732.0, "step": 13561 }, { "epoch": 1.725225798244498, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.847305297851562, "learning_rate": 1e-06, "loss": 0.4689, "mean_token_accuracy": 0.8522243499755859, "num_tokens": 517260423.0, "step": 13562 }, { "epoch": 1.7253530085230886, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.881746292114258, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8567514419555664, "num_tokens": 517300862.0, "step": 13563 }, { "epoch": 1.7254802188016791, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.83949089050293, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8732110261917114, "num_tokens": 517341010.0, "step": 13564 }, { "epoch": 1.7256074290802697, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.870710372924805, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8517127633094788, "num_tokens": 517376853.0, "step": 13565 }, { "epoch": 1.7257346393588602, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.828775405883789, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8636718392372131, "num_tokens": 517421854.0, "step": 13566 }, { "epoch": 1.7258618496374507, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.843863487243652, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8629015684127808, "num_tokens": 517461711.0, "step": 13567 }, { "epoch": 1.7259890599160412, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.911092758178711, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.875072717666626, "num_tokens": 517494755.0, "step": 13568 }, { "epoch": 1.7261162701946318, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.817469596862793, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8598439693450928, "num_tokens": 517539061.0, "step": 13569 }, { "epoch": 1.7262434804732223, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.870518684387207, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8640482425689697, "num_tokens": 517579768.0, "step": 13570 }, { "epoch": 1.7263706907518128, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.871844291687012, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8793009519577026, "num_tokens": 517612006.0, "step": 13571 }, { "epoch": 1.7264979010304033, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.894017219543457, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8682072162628174, "num_tokens": 517647547.0, "step": 13572 }, { "epoch": 1.7266251113089939, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.896035194396973, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8582189083099365, "num_tokens": 517689067.0, "step": 13573 }, { "epoch": 1.7267523215875844, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.869548797607422, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.873416543006897, "num_tokens": 517729967.0, "step": 13574 }, { "epoch": 1.726879531866175, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.839802742004395, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8681886196136475, "num_tokens": 517770427.0, "step": 13575 }, { "epoch": 1.7270067421447655, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.885385513305664, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8671987056732178, "num_tokens": 517804202.0, "step": 13576 }, { "epoch": 1.7271339524233558, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.900047302246094, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8605639934539795, "num_tokens": 517834825.0, "step": 13577 }, { "epoch": 1.7272611627019463, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.86007308959961, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8604834079742432, "num_tokens": 517874600.0, "step": 13578 }, { "epoch": 1.7273883729805368, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.88674259185791, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8587042689323425, "num_tokens": 517913488.0, "step": 13579 }, { "epoch": 1.7275155832591274, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.886679649353027, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.882422924041748, "num_tokens": 517951094.0, "step": 13580 }, { "epoch": 1.7276427935377179, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.849027633666992, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.857883095741272, "num_tokens": 517993811.0, "step": 13581 }, { "epoch": 1.7277700038163084, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.826130867004395, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8704001903533936, "num_tokens": 518032860.0, "step": 13582 }, { "epoch": 1.7278972140948987, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.860776901245117, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8641639947891235, "num_tokens": 518070256.0, "step": 13583 }, { "epoch": 1.7280244243734892, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.91091251373291, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8751817345619202, "num_tokens": 518102828.0, "step": 13584 }, { "epoch": 1.7281516346520798, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.796554565429688, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8820012211799622, "num_tokens": 518146859.0, "step": 13585 }, { "epoch": 1.7282788449306703, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.923782348632812, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8746192455291748, "num_tokens": 518186173.0, "step": 13586 }, { "epoch": 1.7284060552092608, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.908092498779297, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8529085516929626, "num_tokens": 518221903.0, "step": 13587 }, { "epoch": 1.7285332654878514, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.850656509399414, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8586043119430542, "num_tokens": 518258048.0, "step": 13588 }, { "epoch": 1.7286604757664419, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.874462127685547, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8596718311309814, "num_tokens": 518299127.0, "step": 13589 }, { "epoch": 1.7287876860450324, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.882878303527832, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8786315321922302, "num_tokens": 518336520.0, "step": 13590 }, { "epoch": 1.728914896323623, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.88977336883545, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8637101650238037, "num_tokens": 518368484.0, "step": 13591 }, { "epoch": 1.7290421066022135, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.867897987365723, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8667168617248535, "num_tokens": 518408439.0, "step": 13592 }, { "epoch": 1.729169316880804, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.8636474609375, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8481605648994446, "num_tokens": 518447414.0, "step": 13593 }, { "epoch": 1.7292965271593945, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.822778701782227, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8645718097686768, "num_tokens": 518486455.0, "step": 13594 }, { "epoch": 1.729423737437985, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.856857299804688, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8848928213119507, "num_tokens": 518527023.0, "step": 13595 }, { "epoch": 1.7295509477165756, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.863702774047852, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8615953922271729, "num_tokens": 518563285.0, "step": 13596 }, { "epoch": 1.729678157995166, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.879923820495605, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8691213130950928, "num_tokens": 518607610.0, "step": 13597 }, { "epoch": 1.7298053682737566, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.809536933898926, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8775960803031921, "num_tokens": 518649434.0, "step": 13598 }, { "epoch": 1.7299325785523472, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.878911972045898, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8751668930053711, "num_tokens": 518684508.0, "step": 13599 }, { "epoch": 1.7300597888309377, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.891938209533691, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8746525049209595, "num_tokens": 518717883.0, "step": 13600 }, { "epoch": 1.730186999109528, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.872119903564453, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8749175667762756, "num_tokens": 518756158.0, "step": 13601 }, { "epoch": 1.7303142093881185, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.916974067687988, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8675777912139893, "num_tokens": 518793003.0, "step": 13602 }, { "epoch": 1.730441419666709, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.863513946533203, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.862249493598938, "num_tokens": 518838485.0, "step": 13603 }, { "epoch": 1.7305686299452996, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.816028594970703, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8624171018600464, "num_tokens": 518885936.0, "step": 13604 }, { "epoch": 1.7306958402238901, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.891765594482422, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8558090925216675, "num_tokens": 518925684.0, "step": 13605 }, { "epoch": 1.7308230505024806, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.810851097106934, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8709394931793213, "num_tokens": 518966438.0, "step": 13606 }, { "epoch": 1.730950260781071, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.860633850097656, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8704956769943237, "num_tokens": 519013576.0, "step": 13607 }, { "epoch": 1.7310774710596615, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.886526107788086, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8656257390975952, "num_tokens": 519045342.0, "step": 13608 }, { "epoch": 1.731204681338252, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.847816467285156, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8531768321990967, "num_tokens": 519088566.0, "step": 13609 }, { "epoch": 1.7313318916168425, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.892410278320312, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8710882663726807, "num_tokens": 519126516.0, "step": 13610 }, { "epoch": 1.731459101895433, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.832854270935059, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8771061301231384, "num_tokens": 519165431.0, "step": 13611 }, { "epoch": 1.7315863121740236, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.836420059204102, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8834882378578186, "num_tokens": 519200991.0, "step": 13612 }, { "epoch": 1.7317135224526141, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.877979278564453, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8749240636825562, "num_tokens": 519245988.0, "step": 13613 }, { "epoch": 1.7318407327312046, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.815664291381836, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8733305931091309, "num_tokens": 519282595.0, "step": 13614 }, { "epoch": 1.7319679430097952, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.81155776977539, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8620866537094116, "num_tokens": 519319061.0, "step": 13615 }, { "epoch": 1.7320951532883857, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.933157920837402, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8657745718955994, "num_tokens": 519358617.0, "step": 13616 }, { "epoch": 1.7322223635669762, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.908910751342773, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8670578002929688, "num_tokens": 519392461.0, "step": 13617 }, { "epoch": 1.7323495738455668, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.828994750976562, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8757190108299255, "num_tokens": 519432893.0, "step": 13618 }, { "epoch": 1.7324767841241573, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.912757873535156, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8685929775238037, "num_tokens": 519473445.0, "step": 13619 }, { "epoch": 1.7326039944027478, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.893632888793945, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8687061071395874, "num_tokens": 519518883.0, "step": 13620 }, { "epoch": 1.7327312046813383, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.877837181091309, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8732914924621582, "num_tokens": 519550123.0, "step": 13621 }, { "epoch": 1.7328584149599289, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.862650871276855, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8594768643379211, "num_tokens": 519592538.0, "step": 13622 }, { "epoch": 1.7329856252385194, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.843926429748535, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8689444065093994, "num_tokens": 519633466.0, "step": 13623 }, { "epoch": 1.73311283551711, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.920863151550293, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8655115962028503, "num_tokens": 519672569.0, "step": 13624 }, { "epoch": 1.7332400457957005, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.860257148742676, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8741510510444641, "num_tokens": 519707383.0, "step": 13625 }, { "epoch": 1.7333672560742908, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.897115707397461, "learning_rate": 1e-06, "loss": 0.4718, "mean_token_accuracy": 0.8508402705192566, "num_tokens": 519748187.0, "step": 13626 }, { "epoch": 1.7334944663528813, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.857178688049316, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8652029037475586, "num_tokens": 519790215.0, "step": 13627 }, { "epoch": 1.7336216766314718, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.927338600158691, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8600366711616516, "num_tokens": 519827278.0, "step": 13628 }, { "epoch": 1.7337488869100623, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.887319564819336, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8739272356033325, "num_tokens": 519860216.0, "step": 13629 }, { "epoch": 1.7338760971886529, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.84792709350586, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8688835501670837, "num_tokens": 519894442.0, "step": 13630 }, { "epoch": 1.7340033074672434, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.93966293334961, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8679327964782715, "num_tokens": 519932524.0, "step": 13631 }, { "epoch": 1.7341305177458337, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.913086891174316, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8677665591239929, "num_tokens": 519974366.0, "step": 13632 }, { "epoch": 1.7342577280244242, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.921960830688477, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8773342967033386, "num_tokens": 520013983.0, "step": 13633 }, { "epoch": 1.7343849383030148, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.904887199401855, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8715553879737854, "num_tokens": 520047671.0, "step": 13634 }, { "epoch": 1.7345121485816053, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.90303897857666, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.864450216293335, "num_tokens": 520086574.0, "step": 13635 }, { "epoch": 1.7346393588601958, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.897192001342773, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8543143272399902, "num_tokens": 520130613.0, "step": 13636 }, { "epoch": 1.7347665691387864, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.892376899719238, "learning_rate": 1e-06, "loss": 0.3499, "mean_token_accuracy": 0.886941134929657, "num_tokens": 520163763.0, "step": 13637 }, { "epoch": 1.7348937794173769, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.956698417663574, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8662223219871521, "num_tokens": 520198663.0, "step": 13638 }, { "epoch": 1.7350209896959674, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.866189002990723, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8713518381118774, "num_tokens": 520231737.0, "step": 13639 }, { "epoch": 1.735148199974558, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.933506965637207, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8736661672592163, "num_tokens": 520275664.0, "step": 13640 }, { "epoch": 1.7352754102531485, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.90665054321289, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8570787906646729, "num_tokens": 520316100.0, "step": 13641 }, { "epoch": 1.735402620531739, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.90056324005127, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8658445477485657, "num_tokens": 520348167.0, "step": 13642 }, { "epoch": 1.7355298308103295, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.883025169372559, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8547358512878418, "num_tokens": 520385234.0, "step": 13643 }, { "epoch": 1.73565704108892, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.871532440185547, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.8508487939834595, "num_tokens": 520423051.0, "step": 13644 }, { "epoch": 1.7357842513675106, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.939031600952148, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8595185875892639, "num_tokens": 520467202.0, "step": 13645 }, { "epoch": 1.735911461646101, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.827816009521484, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8658957481384277, "num_tokens": 520501434.0, "step": 13646 }, { "epoch": 1.7360386719246916, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.961983680725098, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8655104637145996, "num_tokens": 520542757.0, "step": 13647 }, { "epoch": 1.7361658822032822, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.882539749145508, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8588017225265503, "num_tokens": 520585216.0, "step": 13648 }, { "epoch": 1.7362930924818727, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.933719635009766, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8686789870262146, "num_tokens": 520620066.0, "step": 13649 }, { "epoch": 1.736420302760463, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.908608436584473, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8746360540390015, "num_tokens": 520660917.0, "step": 13650 }, { "epoch": 1.7365475130390535, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.893837928771973, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8603985905647278, "num_tokens": 520704255.0, "step": 13651 }, { "epoch": 1.736674723317644, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.838271141052246, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8547324538230896, "num_tokens": 520743036.0, "step": 13652 }, { "epoch": 1.7368019335962346, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.949578285217285, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8606565594673157, "num_tokens": 520778325.0, "step": 13653 }, { "epoch": 1.736929143874825, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.873916625976562, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8800280690193176, "num_tokens": 520815226.0, "step": 13654 }, { "epoch": 1.7370563541534156, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.961177825927734, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8485721945762634, "num_tokens": 520856606.0, "step": 13655 }, { "epoch": 1.737183564432006, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.923004150390625, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.858257532119751, "num_tokens": 520893493.0, "step": 13656 }, { "epoch": 1.7373107747105965, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.941117286682129, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8754249811172485, "num_tokens": 520930391.0, "step": 13657 }, { "epoch": 1.737437984989187, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.884174346923828, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8681983351707458, "num_tokens": 520970512.0, "step": 13658 }, { "epoch": 1.7375651952677775, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.923576354980469, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8610918521881104, "num_tokens": 521011787.0, "step": 13659 }, { "epoch": 1.737692405546368, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.897136688232422, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8693350553512573, "num_tokens": 521053494.0, "step": 13660 }, { "epoch": 1.7378196158249586, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.975030899047852, "learning_rate": 1e-06, "loss": 0.4565, "mean_token_accuracy": 0.8565690517425537, "num_tokens": 521089291.0, "step": 13661 }, { "epoch": 1.7379468261035491, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.92969799041748, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8795303106307983, "num_tokens": 521128910.0, "step": 13662 }, { "epoch": 1.7380740363821396, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.924283981323242, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.864449679851532, "num_tokens": 521168175.0, "step": 13663 }, { "epoch": 1.7382012466607302, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.902490615844727, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8636654019355774, "num_tokens": 521208020.0, "step": 13664 }, { "epoch": 1.7383284569393207, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.935750961303711, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8795047402381897, "num_tokens": 521249828.0, "step": 13665 }, { "epoch": 1.7384556672179112, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.86789321899414, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8717219829559326, "num_tokens": 521289940.0, "step": 13666 }, { "epoch": 1.7385828774965018, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.917531967163086, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8855640888214111, "num_tokens": 521327426.0, "step": 13667 }, { "epoch": 1.7387100877750923, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.867681503295898, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8578019142150879, "num_tokens": 521372112.0, "step": 13668 }, { "epoch": 1.7388372980536828, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.886617660522461, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8668071031570435, "num_tokens": 521410842.0, "step": 13669 }, { "epoch": 1.7389645083322733, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.89625072479248, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8778379559516907, "num_tokens": 521446854.0, "step": 13670 }, { "epoch": 1.7390917186108639, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.894168853759766, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8579537868499756, "num_tokens": 521486284.0, "step": 13671 }, { "epoch": 1.7392189288894544, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.91897201538086, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8549011945724487, "num_tokens": 521522716.0, "step": 13672 }, { "epoch": 1.739346139168045, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.948627471923828, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.861880898475647, "num_tokens": 521551549.0, "step": 13673 }, { "epoch": 1.7394733494466355, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.911970138549805, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8708451986312866, "num_tokens": 521589552.0, "step": 13674 }, { "epoch": 1.7396005597252258, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.884669303894043, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8800901174545288, "num_tokens": 521627906.0, "step": 13675 }, { "epoch": 1.7397277700038163, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.875113487243652, "learning_rate": 1e-06, "loss": 0.4687, "mean_token_accuracy": 0.8514623045921326, "num_tokens": 521666615.0, "step": 13676 }, { "epoch": 1.7398549802824068, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.866573333740234, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8634587526321411, "num_tokens": 521707742.0, "step": 13677 }, { "epoch": 1.7399821905609973, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.908324241638184, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8576303720474243, "num_tokens": 521746848.0, "step": 13678 }, { "epoch": 1.7401094008395879, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.912392616271973, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8691248893737793, "num_tokens": 521782917.0, "step": 13679 }, { "epoch": 1.7402366111181784, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.889081001281738, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8652700185775757, "num_tokens": 521825609.0, "step": 13680 }, { "epoch": 1.7403638213967687, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.892659187316895, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.881056547164917, "num_tokens": 521862277.0, "step": 13681 }, { "epoch": 1.7404910316753592, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.947522163391113, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8856213688850403, "num_tokens": 521894951.0, "step": 13682 }, { "epoch": 1.7406182419539498, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.859167098999023, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8668104410171509, "num_tokens": 521932158.0, "step": 13683 }, { "epoch": 1.7407454522325403, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.906062126159668, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8740680813789368, "num_tokens": 521971050.0, "step": 13684 }, { "epoch": 1.7408726625111308, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.899165153503418, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8789178133010864, "num_tokens": 522007617.0, "step": 13685 }, { "epoch": 1.7409998727897213, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.90794849395752, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8559328317642212, "num_tokens": 522053484.0, "step": 13686 }, { "epoch": 1.7411270830683119, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.84846019744873, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8642305135726929, "num_tokens": 522088047.0, "step": 13687 }, { "epoch": 1.7412542933469024, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.885509490966797, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8658233880996704, "num_tokens": 522129865.0, "step": 13688 }, { "epoch": 1.741381503625493, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.87237548828125, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8631843328475952, "num_tokens": 522170154.0, "step": 13689 }, { "epoch": 1.7415087139040835, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.86752986907959, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.8920005559921265, "num_tokens": 522204586.0, "step": 13690 }, { "epoch": 1.741635924182674, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.873117446899414, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8714724779129028, "num_tokens": 522241942.0, "step": 13691 }, { "epoch": 1.7417631344612645, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.832125663757324, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8640260696411133, "num_tokens": 522275800.0, "step": 13692 }, { "epoch": 1.741890344739855, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.839326858520508, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8617075085639954, "num_tokens": 522317024.0, "step": 13693 }, { "epoch": 1.7420175550184456, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.863714218139648, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8777716159820557, "num_tokens": 522358510.0, "step": 13694 }, { "epoch": 1.742144765297036, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.823960304260254, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8623153567314148, "num_tokens": 522395960.0, "step": 13695 }, { "epoch": 1.7422719755756266, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.18238639831543, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8764810562133789, "num_tokens": 522428807.0, "step": 13696 }, { "epoch": 1.7423991858542172, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.806726455688477, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8692534565925598, "num_tokens": 522464472.0, "step": 13697 }, { "epoch": 1.7425263961328077, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.832167625427246, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.881585955619812, "num_tokens": 522509522.0, "step": 13698 }, { "epoch": 1.742653606411398, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.86773681640625, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8530255556106567, "num_tokens": 522545581.0, "step": 13699 }, { "epoch": 1.7427808166899885, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.86170768737793, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8718039989471436, "num_tokens": 522582072.0, "step": 13700 }, { "epoch": 1.742908026968579, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.866604804992676, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8715084791183472, "num_tokens": 522618183.0, "step": 13701 }, { "epoch": 1.7430352372471696, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.925005912780762, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.857019305229187, "num_tokens": 522660578.0, "step": 13702 }, { "epoch": 1.74316244752576, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.823868751525879, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8792264461517334, "num_tokens": 522701934.0, "step": 13703 }, { "epoch": 1.7432896578043506, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.871468544006348, "learning_rate": 1e-06, "loss": 0.3595, "mean_token_accuracy": 0.8860237002372742, "num_tokens": 522737404.0, "step": 13704 }, { "epoch": 1.743416868082941, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.941683769226074, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8649730682373047, "num_tokens": 522774188.0, "step": 13705 }, { "epoch": 1.7435440783615315, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.855833053588867, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8769726753234863, "num_tokens": 522810467.0, "step": 13706 }, { "epoch": 1.743671288640122, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.9586181640625, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8577237129211426, "num_tokens": 522845246.0, "step": 13707 }, { "epoch": 1.7437984989187125, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.917922019958496, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.878343939781189, "num_tokens": 522877150.0, "step": 13708 }, { "epoch": 1.743925709197303, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.938158988952637, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8832792639732361, "num_tokens": 522913161.0, "step": 13709 }, { "epoch": 1.7440529194758936, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.867349624633789, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8553006052970886, "num_tokens": 522953464.0, "step": 13710 }, { "epoch": 1.744180129754484, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.898918151855469, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8716837763786316, "num_tokens": 522996874.0, "step": 13711 }, { "epoch": 1.7443073400330746, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.892473220825195, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8715323209762573, "num_tokens": 523031349.0, "step": 13712 }, { "epoch": 1.7444345503116652, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.916557312011719, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8613008260726929, "num_tokens": 523068910.0, "step": 13713 }, { "epoch": 1.7445617605902557, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.894481658935547, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8812201619148254, "num_tokens": 523106179.0, "step": 13714 }, { "epoch": 1.7446889708688462, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.906087875366211, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8699012994766235, "num_tokens": 523140827.0, "step": 13715 }, { "epoch": 1.7448161811474368, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.881230354309082, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.882781982421875, "num_tokens": 523182880.0, "step": 13716 }, { "epoch": 1.7449433914260273, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.90260124206543, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8716796040534973, "num_tokens": 523213059.0, "step": 13717 }, { "epoch": 1.7450706017046178, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.929287910461426, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8679407835006714, "num_tokens": 523254398.0, "step": 13718 }, { "epoch": 1.7451978119832083, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.868557929992676, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8811906576156616, "num_tokens": 523293325.0, "step": 13719 }, { "epoch": 1.7453250222617989, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.856658935546875, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8586896657943726, "num_tokens": 523334402.0, "step": 13720 }, { "epoch": 1.7454522325403894, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.922350883483887, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8730955123901367, "num_tokens": 523376238.0, "step": 13721 }, { "epoch": 1.74557944281898, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.912440299987793, "learning_rate": 1e-06, "loss": 0.3431, "mean_token_accuracy": 0.8908843994140625, "num_tokens": 523417074.0, "step": 13722 }, { "epoch": 1.7457066530975704, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.877284049987793, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8753169775009155, "num_tokens": 523457984.0, "step": 13723 }, { "epoch": 1.7458338633761608, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.89178466796875, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8645085096359253, "num_tokens": 523490071.0, "step": 13724 }, { "epoch": 1.7459610736547513, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.90137004852295, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8628506660461426, "num_tokens": 523530105.0, "step": 13725 }, { "epoch": 1.7460882839333418, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.811493873596191, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8822776675224304, "num_tokens": 523576698.0, "step": 13726 }, { "epoch": 1.7462154942119323, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 12.919723510742188, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8681246042251587, "num_tokens": 523614387.0, "step": 13727 }, { "epoch": 1.7463427044905229, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.906879425048828, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8654123544692993, "num_tokens": 523651149.0, "step": 13728 }, { "epoch": 1.7464699147691134, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.908943176269531, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8819409608840942, "num_tokens": 523688448.0, "step": 13729 }, { "epoch": 1.7465971250477037, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.872579574584961, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8734406232833862, "num_tokens": 523725805.0, "step": 13730 }, { "epoch": 1.7467243353262942, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.88598918914795, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8709825277328491, "num_tokens": 523763564.0, "step": 13731 }, { "epoch": 1.7468515456048848, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.905014038085938, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8640651702880859, "num_tokens": 523804422.0, "step": 13732 }, { "epoch": 1.7469787558834753, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.842184066772461, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.878252387046814, "num_tokens": 523845889.0, "step": 13733 }, { "epoch": 1.7471059661620658, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.898873329162598, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8796874284744263, "num_tokens": 523883623.0, "step": 13734 }, { "epoch": 1.7472331764406563, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.875303268432617, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8755618333816528, "num_tokens": 523917273.0, "step": 13735 }, { "epoch": 1.7473603867192469, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.879007339477539, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8665897846221924, "num_tokens": 523956149.0, "step": 13736 }, { "epoch": 1.7474875969978374, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.86306095123291, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8666239380836487, "num_tokens": 523997229.0, "step": 13737 }, { "epoch": 1.747614807276428, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.84575366973877, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8722738027572632, "num_tokens": 524032921.0, "step": 13738 }, { "epoch": 1.7477420175550185, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.886338233947754, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8706851005554199, "num_tokens": 524065016.0, "step": 13739 }, { "epoch": 1.747869227833609, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.901930809020996, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8579252362251282, "num_tokens": 524107001.0, "step": 13740 }, { "epoch": 1.7479964381121995, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.898664474487305, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8633995056152344, "num_tokens": 524149283.0, "step": 13741 }, { "epoch": 1.74812364839079, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.830244064331055, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8692376613616943, "num_tokens": 524187240.0, "step": 13742 }, { "epoch": 1.7482508586693806, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.846369743347168, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8781899213790894, "num_tokens": 524216793.0, "step": 13743 }, { "epoch": 1.748378068947971, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.87082290649414, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8750777244567871, "num_tokens": 524257629.0, "step": 13744 }, { "epoch": 1.7485052792265616, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.881433486938477, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.88013756275177, "num_tokens": 524291090.0, "step": 13745 }, { "epoch": 1.7486324895051522, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.873692512512207, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8821237087249756, "num_tokens": 524324882.0, "step": 13746 }, { "epoch": 1.7487596997837427, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.880817413330078, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.853107213973999, "num_tokens": 524359422.0, "step": 13747 }, { "epoch": 1.748886910062333, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.837897300720215, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8758182525634766, "num_tokens": 524398383.0, "step": 13748 }, { "epoch": 1.7490141203409235, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.89240837097168, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8652817606925964, "num_tokens": 524437984.0, "step": 13749 }, { "epoch": 1.749141330619514, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.9224271774292, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8751824498176575, "num_tokens": 524471324.0, "step": 13750 }, { "epoch": 1.7492685408981046, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.84803295135498, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8601999282836914, "num_tokens": 524511471.0, "step": 13751 }, { "epoch": 1.749395751176695, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.910357475280762, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8736737966537476, "num_tokens": 524541743.0, "step": 13752 }, { "epoch": 1.7495229614552856, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.946246147155762, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8696118593215942, "num_tokens": 524581470.0, "step": 13753 }, { "epoch": 1.749650171733876, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.89078140258789, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.871924877166748, "num_tokens": 524612722.0, "step": 13754 }, { "epoch": 1.7497773820124665, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.936492919921875, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8775523900985718, "num_tokens": 524647640.0, "step": 13755 }, { "epoch": 1.749904592291057, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.905375480651855, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8627588748931885, "num_tokens": 524682366.0, "step": 13756 }, { "epoch": 1.7500318025696475, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.897162437438965, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8695146441459656, "num_tokens": 524724242.0, "step": 13757 }, { "epoch": 1.750159012848238, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.904740333557129, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8563125133514404, "num_tokens": 524757373.0, "step": 13758 }, { "epoch": 1.7502862231268286, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.888465881347656, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.858893632888794, "num_tokens": 524797704.0, "step": 13759 }, { "epoch": 1.750413433405419, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.864856719970703, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8728611469268799, "num_tokens": 524836710.0, "step": 13760 }, { "epoch": 1.7505406436840096, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.940492630004883, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8673709630966187, "num_tokens": 524870224.0, "step": 13761 }, { "epoch": 1.7506678539626002, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.955162048339844, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8720107078552246, "num_tokens": 524905345.0, "step": 13762 }, { "epoch": 1.7507950642411907, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.88565731048584, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.853226900100708, "num_tokens": 524943893.0, "step": 13763 }, { "epoch": 1.7509222745197812, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.921873092651367, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8684728145599365, "num_tokens": 524982924.0, "step": 13764 }, { "epoch": 1.7510494847983717, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.911824226379395, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8680375814437866, "num_tokens": 525016749.0, "step": 13765 }, { "epoch": 1.7511766950769623, "ewc_loss": 0.03125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.918502807617188, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8711856603622437, "num_tokens": 525049178.0, "step": 13766 }, { "epoch": 1.7513039053555528, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.894225120544434, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8555058836936951, "num_tokens": 525083456.0, "step": 13767 }, { "epoch": 1.7514311156341433, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.876241683959961, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8708569407463074, "num_tokens": 525120159.0, "step": 13768 }, { "epoch": 1.7515583259127339, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.870977401733398, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8748899102210999, "num_tokens": 525159174.0, "step": 13769 }, { "epoch": 1.7516855361913244, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.935182571411133, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8665562868118286, "num_tokens": 525198373.0, "step": 13770 }, { "epoch": 1.751812746469915, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.8370943069458, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8753138780593872, "num_tokens": 525240618.0, "step": 13771 }, { "epoch": 1.7519399567485054, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.92201042175293, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8829150199890137, "num_tokens": 525274899.0, "step": 13772 }, { "epoch": 1.7520671670270958, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.893994331359863, "learning_rate": 1e-06, "loss": 0.4855, "mean_token_accuracy": 0.8495805263519287, "num_tokens": 525310144.0, "step": 13773 }, { "epoch": 1.7521943773056863, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.839134216308594, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8791508078575134, "num_tokens": 525348565.0, "step": 13774 }, { "epoch": 1.7523215875842768, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.910888671875, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8752849698066711, "num_tokens": 525383709.0, "step": 13775 }, { "epoch": 1.7524487978628673, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.829424858093262, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8634324073791504, "num_tokens": 525423983.0, "step": 13776 }, { "epoch": 1.7525760081414579, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.93944263458252, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8760285377502441, "num_tokens": 525457357.0, "step": 13777 }, { "epoch": 1.7527032184200484, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.91260814666748, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8673911094665527, "num_tokens": 525497021.0, "step": 13778 }, { "epoch": 1.7528304286986387, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.910719871520996, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8854678273200989, "num_tokens": 525533592.0, "step": 13779 }, { "epoch": 1.7529576389772292, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.927656173706055, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8615652918815613, "num_tokens": 525572747.0, "step": 13780 }, { "epoch": 1.7530848492558198, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.895092964172363, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8751921057701111, "num_tokens": 525613578.0, "step": 13781 }, { "epoch": 1.7532120595344103, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.94394302368164, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8761056065559387, "num_tokens": 525655047.0, "step": 13782 }, { "epoch": 1.7533392698130008, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.935575485229492, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8666576147079468, "num_tokens": 525689561.0, "step": 13783 }, { "epoch": 1.7534664800915913, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.921914100646973, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8658546209335327, "num_tokens": 525726946.0, "step": 13784 }, { "epoch": 1.7535936903701819, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.95660400390625, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8626928329467773, "num_tokens": 525762089.0, "step": 13785 }, { "epoch": 1.7537209006487724, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.914395332336426, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8748847246170044, "num_tokens": 525796081.0, "step": 13786 }, { "epoch": 1.753848110927363, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.90703010559082, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8586528897285461, "num_tokens": 525829945.0, "step": 13787 }, { "epoch": 1.7539753212059535, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.912109375, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8771508932113647, "num_tokens": 525871289.0, "step": 13788 }, { "epoch": 1.754102531484544, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.870379447937012, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8711140155792236, "num_tokens": 525908455.0, "step": 13789 }, { "epoch": 1.7542297417631345, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.846528053283691, "learning_rate": 1e-06, "loss": 0.3412, "mean_token_accuracy": 0.8906722068786621, "num_tokens": 525943734.0, "step": 13790 }, { "epoch": 1.754356952041725, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.896740913391113, "learning_rate": 1e-06, "loss": 0.3401, "mean_token_accuracy": 0.8910974264144897, "num_tokens": 525988070.0, "step": 13791 }, { "epoch": 1.7544841623203156, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.837316513061523, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8840246200561523, "num_tokens": 526027940.0, "step": 13792 }, { "epoch": 1.754611372598906, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.885849952697754, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.879152238368988, "num_tokens": 526061357.0, "step": 13793 }, { "epoch": 1.7547385828774966, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.90274715423584, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.872117817401886, "num_tokens": 526102157.0, "step": 13794 }, { "epoch": 1.7548657931560872, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.90020751953125, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8618297576904297, "num_tokens": 526137241.0, "step": 13795 }, { "epoch": 1.7549930034346777, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.896933555603027, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8618640899658203, "num_tokens": 526169079.0, "step": 13796 }, { "epoch": 1.755120213713268, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.891441345214844, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8747986555099487, "num_tokens": 526212212.0, "step": 13797 }, { "epoch": 1.7552474239918585, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.854825019836426, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8773155212402344, "num_tokens": 526250380.0, "step": 13798 }, { "epoch": 1.755374634270449, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.917908668518066, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8771242499351501, "num_tokens": 526292525.0, "step": 13799 }, { "epoch": 1.7555018445490396, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.895121574401855, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8710985779762268, "num_tokens": 526330427.0, "step": 13800 }, { "epoch": 1.75562905482763, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.910856246948242, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8724610209465027, "num_tokens": 526362954.0, "step": 13801 }, { "epoch": 1.7557562651062206, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.88125228881836, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.864030122756958, "num_tokens": 526394310.0, "step": 13802 }, { "epoch": 1.755883475384811, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.880626678466797, "learning_rate": 1e-06, "loss": 0.4656, "mean_token_accuracy": 0.8505197167396545, "num_tokens": 526438507.0, "step": 13803 }, { "epoch": 1.7560106856634015, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.906150817871094, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.866605281829834, "num_tokens": 526475124.0, "step": 13804 }, { "epoch": 1.756137895941992, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.88438606262207, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8790029287338257, "num_tokens": 526514936.0, "step": 13805 }, { "epoch": 1.7562651062205825, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.867414474487305, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8642085194587708, "num_tokens": 526548649.0, "step": 13806 }, { "epoch": 1.756392316499173, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.938485145568848, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8533235192298889, "num_tokens": 526589028.0, "step": 13807 }, { "epoch": 1.7565195267777636, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.848631858825684, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8704351186752319, "num_tokens": 526626327.0, "step": 13808 }, { "epoch": 1.756646737056354, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.913898468017578, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8666952252388, "num_tokens": 526672264.0, "step": 13809 }, { "epoch": 1.7567739473349446, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.915799140930176, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.86342453956604, "num_tokens": 526709979.0, "step": 13810 }, { "epoch": 1.7569011576135352, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.903103828430176, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8823745250701904, "num_tokens": 526745002.0, "step": 13811 }, { "epoch": 1.7570283678921257, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.883380889892578, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8851243853569031, "num_tokens": 526785684.0, "step": 13812 }, { "epoch": 1.7571555781707162, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.84516716003418, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8875375390052795, "num_tokens": 526820246.0, "step": 13813 }, { "epoch": 1.7572827884493067, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.909095764160156, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8742581605911255, "num_tokens": 526854815.0, "step": 13814 }, { "epoch": 1.7574099987278973, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.839051246643066, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8755652904510498, "num_tokens": 526896336.0, "step": 13815 }, { "epoch": 1.7575372090064878, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.887696266174316, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8818208575248718, "num_tokens": 526939586.0, "step": 13816 }, { "epoch": 1.7576644192850783, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.866244316101074, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8752966523170471, "num_tokens": 526975027.0, "step": 13817 }, { "epoch": 1.7577916295636689, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.900214195251465, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8718812465667725, "num_tokens": 527017820.0, "step": 13818 }, { "epoch": 1.7579188398422594, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.92870044708252, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8807877898216248, "num_tokens": 527051647.0, "step": 13819 }, { "epoch": 1.75804605012085, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.922099113464355, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8750631213188171, "num_tokens": 527089590.0, "step": 13820 }, { "epoch": 1.7581732603994404, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.907671928405762, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8679472804069519, "num_tokens": 527126696.0, "step": 13821 }, { "epoch": 1.7583004706780307, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.940783500671387, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8622825145721436, "num_tokens": 527164823.0, "step": 13822 }, { "epoch": 1.7584276809566213, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.934158325195312, "learning_rate": 1e-06, "loss": 0.4633, "mean_token_accuracy": 0.8540326952934265, "num_tokens": 527208032.0, "step": 13823 }, { "epoch": 1.7585548912352118, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.892045021057129, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8717015981674194, "num_tokens": 527242773.0, "step": 13824 }, { "epoch": 1.7586821015138023, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.89552116394043, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8735626935958862, "num_tokens": 527277968.0, "step": 13825 }, { "epoch": 1.7588093117923929, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.918845176696777, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8840572834014893, "num_tokens": 527311077.0, "step": 13826 }, { "epoch": 1.7589365220709834, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.94299030303955, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8760020732879639, "num_tokens": 527340543.0, "step": 13827 }, { "epoch": 1.7590637323495737, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.87687873840332, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8684093952178955, "num_tokens": 527379596.0, "step": 13828 }, { "epoch": 1.7591909426281642, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.935111045837402, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8704533576965332, "num_tokens": 527417873.0, "step": 13829 }, { "epoch": 1.7593181529067548, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.97374153137207, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.872346043586731, "num_tokens": 527458607.0, "step": 13830 }, { "epoch": 1.7594453631853453, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.895679473876953, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8618776798248291, "num_tokens": 527493299.0, "step": 13831 }, { "epoch": 1.7595725734639358, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.908612251281738, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8580868244171143, "num_tokens": 527536065.0, "step": 13832 }, { "epoch": 1.7596997837425263, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.879071235656738, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8556247353553772, "num_tokens": 527576477.0, "step": 13833 }, { "epoch": 1.7598269940211169, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.936351776123047, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8678134679794312, "num_tokens": 527611777.0, "step": 13834 }, { "epoch": 1.7599542042997074, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.954642295837402, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8880983591079712, "num_tokens": 527650236.0, "step": 13835 }, { "epoch": 1.760081414578298, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.863396644592285, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8774723410606384, "num_tokens": 527685486.0, "step": 13836 }, { "epoch": 1.7602086248568884, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.95209789276123, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8577082753181458, "num_tokens": 527725386.0, "step": 13837 }, { "epoch": 1.760335835135479, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.003957748413086, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8579788208007812, "num_tokens": 527756213.0, "step": 13838 }, { "epoch": 1.7604630454140695, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.905877113342285, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8751802444458008, "num_tokens": 527793884.0, "step": 13839 }, { "epoch": 1.76059025569266, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.94394302368164, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8679698705673218, "num_tokens": 527826428.0, "step": 13840 }, { "epoch": 1.7607174659712506, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.952877044677734, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8626231551170349, "num_tokens": 527866691.0, "step": 13841 }, { "epoch": 1.760844676249841, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.930313110351562, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8789161443710327, "num_tokens": 527909482.0, "step": 13842 }, { "epoch": 1.7609718865284316, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.930365562438965, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8587774038314819, "num_tokens": 527947818.0, "step": 13843 }, { "epoch": 1.7610990968070221, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.891580581665039, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8742142915725708, "num_tokens": 527986760.0, "step": 13844 }, { "epoch": 1.7612263070856127, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.920476913452148, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8636338710784912, "num_tokens": 528024742.0, "step": 13845 }, { "epoch": 1.761353517364203, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.938827514648438, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8567005395889282, "num_tokens": 528060781.0, "step": 13846 }, { "epoch": 1.7614807276427935, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.872748374938965, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8720170259475708, "num_tokens": 528099051.0, "step": 13847 }, { "epoch": 1.761607937921384, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.904634475708008, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8639451265335083, "num_tokens": 528138325.0, "step": 13848 }, { "epoch": 1.7617351481999746, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.899654388427734, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8610470294952393, "num_tokens": 528177997.0, "step": 13849 }, { "epoch": 1.761862358478565, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.88593578338623, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8668729066848755, "num_tokens": 528216360.0, "step": 13850 }, { "epoch": 1.7619895687571556, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.92857837677002, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8849866390228271, "num_tokens": 528251100.0, "step": 13851 }, { "epoch": 1.762116779035746, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.93184757232666, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8744177222251892, "num_tokens": 528286298.0, "step": 13852 }, { "epoch": 1.7622439893143365, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.8688325881958, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8863986730575562, "num_tokens": 528322970.0, "step": 13853 }, { "epoch": 1.762371199592927, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.9405517578125, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8863869309425354, "num_tokens": 528355191.0, "step": 13854 }, { "epoch": 1.7624984098715175, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.912186622619629, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8694564700126648, "num_tokens": 528388975.0, "step": 13855 }, { "epoch": 1.762625620150108, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.882497787475586, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8749959468841553, "num_tokens": 528432139.0, "step": 13856 }, { "epoch": 1.7627528304286986, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.862482070922852, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8798050880432129, "num_tokens": 528468315.0, "step": 13857 }, { "epoch": 1.762880040707289, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.920708656311035, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8597962856292725, "num_tokens": 528510368.0, "step": 13858 }, { "epoch": 1.7630072509858796, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.893589973449707, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8728072047233582, "num_tokens": 528545637.0, "step": 13859 }, { "epoch": 1.7631344612644702, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.9442138671875, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8754344582557678, "num_tokens": 528586267.0, "step": 13860 }, { "epoch": 1.7632616715430607, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.88782787322998, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.873854398727417, "num_tokens": 528629640.0, "step": 13861 }, { "epoch": 1.7633888818216512, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.000970840454102, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8560886979103088, "num_tokens": 528665373.0, "step": 13862 }, { "epoch": 1.7635160921002417, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.942663192749023, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8837852478027344, "num_tokens": 528703426.0, "step": 13863 }, { "epoch": 1.7636433023788323, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.918094635009766, "learning_rate": 1e-06, "loss": 0.4724, "mean_token_accuracy": 0.8520880937576294, "num_tokens": 528744886.0, "step": 13864 }, { "epoch": 1.7637705126574228, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.959007263183594, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8827416896820068, "num_tokens": 528779773.0, "step": 13865 }, { "epoch": 1.7638977229360133, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.920368194580078, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8727151155471802, "num_tokens": 528821878.0, "step": 13866 }, { "epoch": 1.7640249332146039, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.95705795288086, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8726502656936646, "num_tokens": 528858056.0, "step": 13867 }, { "epoch": 1.7641521434931944, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.905817031860352, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8676246404647827, "num_tokens": 528903116.0, "step": 13868 }, { "epoch": 1.764279353771785, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 12.882730484008789, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.86464524269104, "num_tokens": 528941561.0, "step": 13869 }, { "epoch": 1.7644065640503754, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.970026969909668, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.8874069452285767, "num_tokens": 528976215.0, "step": 13870 }, { "epoch": 1.7645337743289657, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.887490272521973, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8765378594398499, "num_tokens": 529019280.0, "step": 13871 }, { "epoch": 1.7646609846075563, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.904938697814941, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8743043541908264, "num_tokens": 529060733.0, "step": 13872 }, { "epoch": 1.7647881948861468, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.941646575927734, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8733973503112793, "num_tokens": 529094948.0, "step": 13873 }, { "epoch": 1.7649154051647373, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.929553985595703, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8817850351333618, "num_tokens": 529131374.0, "step": 13874 }, { "epoch": 1.7650426154433279, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.945734024047852, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8675689101219177, "num_tokens": 529169133.0, "step": 13875 }, { "epoch": 1.7651698257219184, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.985135078430176, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8686879873275757, "num_tokens": 529206467.0, "step": 13876 }, { "epoch": 1.7652970360005087, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.906167030334473, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.861299991607666, "num_tokens": 529244891.0, "step": 13877 }, { "epoch": 1.7654242462790992, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.99108600616455, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8723405599594116, "num_tokens": 529277384.0, "step": 13878 }, { "epoch": 1.7655514565576897, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.901869773864746, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8748844861984253, "num_tokens": 529322530.0, "step": 13879 }, { "epoch": 1.7656786668362803, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.883411407470703, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8658877611160278, "num_tokens": 529356345.0, "step": 13880 }, { "epoch": 1.7658058771148708, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.94160270690918, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8727972507476807, "num_tokens": 529391877.0, "step": 13881 }, { "epoch": 1.7659330873934613, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.978131294250488, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8782660365104675, "num_tokens": 529425460.0, "step": 13882 }, { "epoch": 1.7660602976720519, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.89706039428711, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8606091737747192, "num_tokens": 529459660.0, "step": 13883 }, { "epoch": 1.7661875079506424, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.909239768981934, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8630964159965515, "num_tokens": 529501839.0, "step": 13884 }, { "epoch": 1.766314718229233, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.930548667907715, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8680643439292908, "num_tokens": 529540828.0, "step": 13885 }, { "epoch": 1.7664419285078234, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.913658142089844, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8704208135604858, "num_tokens": 529581578.0, "step": 13886 }, { "epoch": 1.766569138786414, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.902544975280762, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8707973957061768, "num_tokens": 529624363.0, "step": 13887 }, { "epoch": 1.7666963490650045, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.886248588562012, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8773242831230164, "num_tokens": 529665286.0, "step": 13888 }, { "epoch": 1.766823559343595, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.905489921569824, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8668419122695923, "num_tokens": 529704150.0, "step": 13889 }, { "epoch": 1.7669507696221856, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.902790069580078, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8727090358734131, "num_tokens": 529742110.0, "step": 13890 }, { "epoch": 1.767077979900776, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.91767692565918, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8695029020309448, "num_tokens": 529776158.0, "step": 13891 }, { "epoch": 1.7672051901793666, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.947805404663086, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.848250150680542, "num_tokens": 529814002.0, "step": 13892 }, { "epoch": 1.7673324004579571, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.889813423156738, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8733887076377869, "num_tokens": 529846696.0, "step": 13893 }, { "epoch": 1.7674596107365477, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.971879005432129, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.867868185043335, "num_tokens": 529881557.0, "step": 13894 }, { "epoch": 1.767586821015138, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.87398624420166, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.8581010103225708, "num_tokens": 529924337.0, "step": 13895 }, { "epoch": 1.7677140312937285, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.983500480651855, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8695553541183472, "num_tokens": 529967195.0, "step": 13896 }, { "epoch": 1.767841241572319, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.989459991455078, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8510078191757202, "num_tokens": 530003454.0, "step": 13897 }, { "epoch": 1.7679684518509096, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.950078010559082, "learning_rate": 1e-06, "loss": 0.4833, "mean_token_accuracy": 0.8438627123832703, "num_tokens": 530042577.0, "step": 13898 }, { "epoch": 1.7680956621295, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.938437461853027, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8554480075836182, "num_tokens": 530084903.0, "step": 13899 }, { "epoch": 1.7682228724080906, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.949557304382324, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8650133609771729, "num_tokens": 530117598.0, "step": 13900 }, { "epoch": 1.768350082686681, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.918656349182129, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8733579516410828, "num_tokens": 530151891.0, "step": 13901 }, { "epoch": 1.7684772929652715, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.9131498336792, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8612030744552612, "num_tokens": 530191263.0, "step": 13902 }, { "epoch": 1.768604503243862, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.902217864990234, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8656361103057861, "num_tokens": 530231156.0, "step": 13903 }, { "epoch": 1.7687317135224525, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.930533409118652, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8837167024612427, "num_tokens": 530270089.0, "step": 13904 }, { "epoch": 1.768858923801043, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.917616844177246, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8741565942764282, "num_tokens": 530309808.0, "step": 13905 }, { "epoch": 1.7689861340796336, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.959663391113281, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8587589859962463, "num_tokens": 530357723.0, "step": 13906 }, { "epoch": 1.769113344358224, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.897815704345703, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8568258285522461, "num_tokens": 530389586.0, "step": 13907 }, { "epoch": 1.7692405546368146, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.930946350097656, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8601410388946533, "num_tokens": 530430356.0, "step": 13908 }, { "epoch": 1.7693677649154052, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.924864768981934, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8742737770080566, "num_tokens": 530472859.0, "step": 13909 }, { "epoch": 1.7694949751939957, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.893939018249512, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8573037385940552, "num_tokens": 530515422.0, "step": 13910 }, { "epoch": 1.7696221854725862, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.938937187194824, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8668689727783203, "num_tokens": 530553842.0, "step": 13911 }, { "epoch": 1.7697493957511767, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.91040325164795, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8687714338302612, "num_tokens": 530592593.0, "step": 13912 }, { "epoch": 1.7698766060297673, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.853422164916992, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.863018810749054, "num_tokens": 530632309.0, "step": 13913 }, { "epoch": 1.7700038163083578, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.962430953979492, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.865543782711029, "num_tokens": 530667415.0, "step": 13914 }, { "epoch": 1.7701310265869483, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.940796852111816, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.867019772529602, "num_tokens": 530705736.0, "step": 13915 }, { "epoch": 1.7702582368655388, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.867623329162598, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.878001868724823, "num_tokens": 530744931.0, "step": 13916 }, { "epoch": 1.7703854471441294, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.938193321228027, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8806353211402893, "num_tokens": 530784700.0, "step": 13917 }, { "epoch": 1.77051265742272, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.912235260009766, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8852088451385498, "num_tokens": 530821097.0, "step": 13918 }, { "epoch": 1.7706398677013102, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.931232452392578, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8638293743133545, "num_tokens": 530860444.0, "step": 13919 }, { "epoch": 1.7707670779799007, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.919116020202637, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8659437894821167, "num_tokens": 530907304.0, "step": 13920 }, { "epoch": 1.7708942882584913, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.966766357421875, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8615925312042236, "num_tokens": 530945692.0, "step": 13921 }, { "epoch": 1.7710214985370818, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.955267906188965, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8792736530303955, "num_tokens": 530986695.0, "step": 13922 }, { "epoch": 1.7711487088156723, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.912983894348145, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8797893524169922, "num_tokens": 531022656.0, "step": 13923 }, { "epoch": 1.7712759190942629, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.96109390258789, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8785028457641602, "num_tokens": 531056896.0, "step": 13924 }, { "epoch": 1.7714031293728534, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.912175178527832, "learning_rate": 1e-06, "loss": 0.4929, "mean_token_accuracy": 0.8436121940612793, "num_tokens": 531097286.0, "step": 13925 }, { "epoch": 1.7715303396514437, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.92712116241455, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.877012312412262, "num_tokens": 531136430.0, "step": 13926 }, { "epoch": 1.7716575499300342, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.91822624206543, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8619943857192993, "num_tokens": 531176698.0, "step": 13927 }, { "epoch": 1.7717847602086247, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.910735130310059, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8688327074050903, "num_tokens": 531214343.0, "step": 13928 }, { "epoch": 1.7719119704872153, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.901972770690918, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8818591833114624, "num_tokens": 531252325.0, "step": 13929 }, { "epoch": 1.7720391807658058, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.950841903686523, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8577022552490234, "num_tokens": 531290378.0, "step": 13930 }, { "epoch": 1.7721663910443963, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.92667293548584, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8600820302963257, "num_tokens": 531330406.0, "step": 13931 }, { "epoch": 1.7722936013229869, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.97425651550293, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8703137040138245, "num_tokens": 531367425.0, "step": 13932 }, { "epoch": 1.7724208116015774, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.919116973876953, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.8904514908790588, "num_tokens": 531406465.0, "step": 13933 }, { "epoch": 1.772548021880168, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.979567527770996, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.859253466129303, "num_tokens": 531445356.0, "step": 13934 }, { "epoch": 1.7726752321587584, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.9271879196167, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8666694164276123, "num_tokens": 531484737.0, "step": 13935 }, { "epoch": 1.772802442437349, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.973694801330566, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8656610250473022, "num_tokens": 531523453.0, "step": 13936 }, { "epoch": 1.7729296527159395, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.915202140808105, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.875818133354187, "num_tokens": 531559627.0, "step": 13937 }, { "epoch": 1.77305686299453, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.897745132446289, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8743650317192078, "num_tokens": 531597677.0, "step": 13938 }, { "epoch": 1.7731840732731206, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.792587280273438, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8779155015945435, "num_tokens": 531642753.0, "step": 13939 }, { "epoch": 1.773311283551711, "ewc_loss": 0.03125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 12.941858291625977, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8619111776351929, "num_tokens": 531688249.0, "step": 13940 }, { "epoch": 1.7734384938303016, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.87492561340332, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8833061456680298, "num_tokens": 531727994.0, "step": 13941 }, { "epoch": 1.7735657041088921, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.930292129516602, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8730811476707458, "num_tokens": 531764428.0, "step": 13942 }, { "epoch": 1.7736929143874827, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.873560905456543, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8646081686019897, "num_tokens": 531806604.0, "step": 13943 }, { "epoch": 1.773820124666073, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.93057918548584, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8874056339263916, "num_tokens": 531845504.0, "step": 13944 }, { "epoch": 1.7739473349446635, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.997753143310547, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8789742588996887, "num_tokens": 531880962.0, "step": 13945 }, { "epoch": 1.774074545223254, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.929311752319336, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8670863509178162, "num_tokens": 531918182.0, "step": 13946 }, { "epoch": 1.7742017555018446, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.912519454956055, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8706597089767456, "num_tokens": 531958340.0, "step": 13947 }, { "epoch": 1.774328965780435, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.91441535949707, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8582752346992493, "num_tokens": 531996951.0, "step": 13948 }, { "epoch": 1.7744561760590256, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.954214096069336, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8734688758850098, "num_tokens": 532036963.0, "step": 13949 }, { "epoch": 1.774583386337616, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.941227912902832, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8638548851013184, "num_tokens": 532070937.0, "step": 13950 }, { "epoch": 1.7747105966162064, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.920843124389648, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8560358285903931, "num_tokens": 532107866.0, "step": 13951 }, { "epoch": 1.774837806894797, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.938014030456543, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8620568513870239, "num_tokens": 532151485.0, "step": 13952 }, { "epoch": 1.7749650171733875, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.870402336120605, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.875949501991272, "num_tokens": 532191209.0, "step": 13953 }, { "epoch": 1.775092227451978, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.922513008117676, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8644157648086548, "num_tokens": 532232219.0, "step": 13954 }, { "epoch": 1.7752194377305686, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.93348503112793, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8569350838661194, "num_tokens": 532265330.0, "step": 13955 }, { "epoch": 1.775346648009159, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 13.041621208190918, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8681614995002747, "num_tokens": 532295964.0, "step": 13956 }, { "epoch": 1.7754738582877496, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.874788284301758, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8727155923843384, "num_tokens": 532333804.0, "step": 13957 }, { "epoch": 1.7756010685663401, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.941431999206543, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8736729025840759, "num_tokens": 532373088.0, "step": 13958 }, { "epoch": 1.7757282788449307, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.919581413269043, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8673463463783264, "num_tokens": 532409473.0, "step": 13959 }, { "epoch": 1.7758554891235212, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.935707092285156, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8797857761383057, "num_tokens": 532447883.0, "step": 13960 }, { "epoch": 1.7759826994021117, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.832398414611816, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8745666742324829, "num_tokens": 532485783.0, "step": 13961 }, { "epoch": 1.7761099096807023, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.044836044311523, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.87519371509552, "num_tokens": 532511082.0, "step": 13962 }, { "epoch": 1.7762371199592928, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.86904525756836, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8855662941932678, "num_tokens": 532555516.0, "step": 13963 }, { "epoch": 1.7763643302378833, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.921975135803223, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8540270328521729, "num_tokens": 532599200.0, "step": 13964 }, { "epoch": 1.7764915405164738, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.968268394470215, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8793570399284363, "num_tokens": 532636335.0, "step": 13965 }, { "epoch": 1.7766187507950644, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.969254493713379, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.881356954574585, "num_tokens": 532678337.0, "step": 13966 }, { "epoch": 1.776745961073655, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.903684616088867, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8658372759819031, "num_tokens": 532715051.0, "step": 13967 }, { "epoch": 1.7768731713522452, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.948159217834473, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8801544904708862, "num_tokens": 532750836.0, "step": 13968 }, { "epoch": 1.7770003816308357, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.884597778320312, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8750466108322144, "num_tokens": 532790331.0, "step": 13969 }, { "epoch": 1.7771275919094263, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.922691345214844, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.865831732749939, "num_tokens": 532830169.0, "step": 13970 }, { "epoch": 1.7772548021880168, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.932092666625977, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8623617887496948, "num_tokens": 532865957.0, "step": 13971 }, { "epoch": 1.7773820124666073, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.867575645446777, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8660587072372437, "num_tokens": 532906112.0, "step": 13972 }, { "epoch": 1.7775092227451978, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.87131404876709, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.875794529914856, "num_tokens": 532939676.0, "step": 13973 }, { "epoch": 1.7776364330237884, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.942955017089844, "learning_rate": 1e-06, "loss": 0.4629, "mean_token_accuracy": 0.8514317274093628, "num_tokens": 532977668.0, "step": 13974 }, { "epoch": 1.7777636433023787, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.910696029663086, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8668338060379028, "num_tokens": 533013862.0, "step": 13975 }, { "epoch": 1.7778908535809692, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.893049240112305, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8583455681800842, "num_tokens": 533051222.0, "step": 13976 }, { "epoch": 1.7780180638595597, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.929055213928223, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8672119379043579, "num_tokens": 533090182.0, "step": 13977 }, { "epoch": 1.7781452741381503, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.871936798095703, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8825601935386658, "num_tokens": 533129600.0, "step": 13978 }, { "epoch": 1.7782724844167408, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.969352722167969, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8689223527908325, "num_tokens": 533172351.0, "step": 13979 }, { "epoch": 1.7783996946953313, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.890104293823242, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8673069477081299, "num_tokens": 533211583.0, "step": 13980 }, { "epoch": 1.7785269049739219, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.931488990783691, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.875556468963623, "num_tokens": 533249960.0, "step": 13981 }, { "epoch": 1.7786541152525124, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.888550758361816, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8765457272529602, "num_tokens": 533287328.0, "step": 13982 }, { "epoch": 1.778781325531103, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.937936782836914, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8699969053268433, "num_tokens": 533327772.0, "step": 13983 }, { "epoch": 1.7789085358096934, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.905682563781738, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8631489872932434, "num_tokens": 533372004.0, "step": 13984 }, { "epoch": 1.779035746088284, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.94720458984375, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8680602312088013, "num_tokens": 533409502.0, "step": 13985 }, { "epoch": 1.7791629563668745, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.033878326416016, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.857744574546814, "num_tokens": 533451160.0, "step": 13986 }, { "epoch": 1.779290166645465, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.930625915527344, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8777239322662354, "num_tokens": 533480413.0, "step": 13987 }, { "epoch": 1.7794173769240555, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.915261268615723, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8577498197555542, "num_tokens": 533518826.0, "step": 13988 }, { "epoch": 1.779544587202646, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.940862655639648, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8586000800132751, "num_tokens": 533563760.0, "step": 13989 }, { "epoch": 1.7796717974812366, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.957353591918945, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8705077171325684, "num_tokens": 533601499.0, "step": 13990 }, { "epoch": 1.7797990077598271, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.960260391235352, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.872861385345459, "num_tokens": 533640757.0, "step": 13991 }, { "epoch": 1.7799262180384177, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.925378799438477, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8716350793838501, "num_tokens": 533678047.0, "step": 13992 }, { "epoch": 1.780053428317008, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.911593437194824, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8696963787078857, "num_tokens": 533720903.0, "step": 13993 }, { "epoch": 1.7801806385955985, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.921602249145508, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8708842396736145, "num_tokens": 533759720.0, "step": 13994 }, { "epoch": 1.780307848874189, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.904603958129883, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8580896258354187, "num_tokens": 533796122.0, "step": 13995 }, { "epoch": 1.7804350591527796, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.956561088562012, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8755348920822144, "num_tokens": 533836326.0, "step": 13996 }, { "epoch": 1.78056226943137, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.946700096130371, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8804793953895569, "num_tokens": 533867874.0, "step": 13997 }, { "epoch": 1.7806894797099606, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.923158645629883, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8820988535881042, "num_tokens": 533902877.0, "step": 13998 }, { "epoch": 1.780816689988551, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.981364250183105, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8730427026748657, "num_tokens": 533937440.0, "step": 13999 }, { "epoch": 1.7809439002671414, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.974162101745605, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8794123530387878, "num_tokens": 533972308.0, "step": 14000 }, { "epoch": 1.781071110545732, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.953822135925293, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8568544983863831, "num_tokens": 534011913.0, "step": 14001 }, { "epoch": 1.7811983208243225, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.903885841369629, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.880631685256958, "num_tokens": 534057235.0, "step": 14002 }, { "epoch": 1.781325531102913, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.007652282714844, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8700300455093384, "num_tokens": 534092850.0, "step": 14003 }, { "epoch": 1.7814527413815036, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.879193305969238, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8745291233062744, "num_tokens": 534135310.0, "step": 14004 }, { "epoch": 1.781579951660094, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.997483253479004, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8688740730285645, "num_tokens": 534176008.0, "step": 14005 }, { "epoch": 1.7817071619386846, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.013689041137695, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8699153065681458, "num_tokens": 534208217.0, "step": 14006 }, { "epoch": 1.7818343722172751, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.880593299865723, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.874231219291687, "num_tokens": 534247564.0, "step": 14007 }, { "epoch": 1.7819615824958657, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.98252010345459, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8509780764579773, "num_tokens": 534281144.0, "step": 14008 }, { "epoch": 1.7820887927744562, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.998077392578125, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8590333461761475, "num_tokens": 534317893.0, "step": 14009 }, { "epoch": 1.7822160030530467, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.9158935546875, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8679572343826294, "num_tokens": 534356116.0, "step": 14010 }, { "epoch": 1.7823432133316373, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.976656913757324, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8441191911697388, "num_tokens": 534390499.0, "step": 14011 }, { "epoch": 1.7824704236102278, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.01158332824707, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8646376132965088, "num_tokens": 534424242.0, "step": 14012 }, { "epoch": 1.7825976338888183, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.982501029968262, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8609294891357422, "num_tokens": 534467465.0, "step": 14013 }, { "epoch": 1.7827248441674088, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.955272674560547, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8635249137878418, "num_tokens": 534502959.0, "step": 14014 }, { "epoch": 1.7828520544459994, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.91076946258545, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8704830408096313, "num_tokens": 534549888.0, "step": 14015 }, { "epoch": 1.78297926472459, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.021344184875488, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8718302845954895, "num_tokens": 534591686.0, "step": 14016 }, { "epoch": 1.7831064750031802, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.938445091247559, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.885260820388794, "num_tokens": 534628584.0, "step": 14017 }, { "epoch": 1.7832336852817707, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.98460578918457, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.859818696975708, "num_tokens": 534665684.0, "step": 14018 }, { "epoch": 1.7833608955603613, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.9662446975708, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8724261522293091, "num_tokens": 534701604.0, "step": 14019 }, { "epoch": 1.7834881058389518, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.941842079162598, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8533911108970642, "num_tokens": 534741765.0, "step": 14020 }, { "epoch": 1.7836153161175423, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.878323554992676, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8787709474563599, "num_tokens": 534780331.0, "step": 14021 }, { "epoch": 1.7837425263961328, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.991312026977539, "learning_rate": 1e-06, "loss": 0.4621, "mean_token_accuracy": 0.8577386736869812, "num_tokens": 534816254.0, "step": 14022 }, { "epoch": 1.7838697366747234, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.945772171020508, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8686195015907288, "num_tokens": 534851367.0, "step": 14023 }, { "epoch": 1.7839969469533137, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.91395378112793, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8673433065414429, "num_tokens": 534889077.0, "step": 14024 }, { "epoch": 1.7841241572319042, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.924832344055176, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8584237098693848, "num_tokens": 534924249.0, "step": 14025 }, { "epoch": 1.7842513675104947, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.94767951965332, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8539184331893921, "num_tokens": 534966370.0, "step": 14026 }, { "epoch": 1.7843785777890853, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.92802619934082, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8753843307495117, "num_tokens": 535005869.0, "step": 14027 }, { "epoch": 1.7845057880676758, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.933121681213379, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8666568994522095, "num_tokens": 535042057.0, "step": 14028 }, { "epoch": 1.7846329983462663, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.99137020111084, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8642404079437256, "num_tokens": 535074410.0, "step": 14029 }, { "epoch": 1.7847602086248568, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.909150123596191, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8594106435775757, "num_tokens": 535120356.0, "step": 14030 }, { "epoch": 1.7848874189034474, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.00356388092041, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8591694235801697, "num_tokens": 535156939.0, "step": 14031 }, { "epoch": 1.785014629182038, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.929071426391602, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8712942600250244, "num_tokens": 535195553.0, "step": 14032 }, { "epoch": 1.7851418394606284, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.876608848571777, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.873586893081665, "num_tokens": 535235692.0, "step": 14033 }, { "epoch": 1.785269049739219, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.966262817382812, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8673976063728333, "num_tokens": 535269265.0, "step": 14034 }, { "epoch": 1.7853962600178095, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.938589096069336, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.883764386177063, "num_tokens": 535303825.0, "step": 14035 }, { "epoch": 1.7855234702964, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.948854446411133, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8639694452285767, "num_tokens": 535338802.0, "step": 14036 }, { "epoch": 1.7856506805749905, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.8839111328125, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8764147162437439, "num_tokens": 535376592.0, "step": 14037 }, { "epoch": 1.785777890853581, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.967449188232422, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8653547763824463, "num_tokens": 535416528.0, "step": 14038 }, { "epoch": 1.7859051011321716, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.965099334716797, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8651697635650635, "num_tokens": 535450892.0, "step": 14039 }, { "epoch": 1.7860323114107621, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.92594051361084, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.871198296546936, "num_tokens": 535488198.0, "step": 14040 }, { "epoch": 1.7861595216893527, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.930683135986328, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8670734763145447, "num_tokens": 535526020.0, "step": 14041 }, { "epoch": 1.786286731967943, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.973877906799316, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8741703033447266, "num_tokens": 535560400.0, "step": 14042 }, { "epoch": 1.7864139422465335, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.930038452148438, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8631024360656738, "num_tokens": 535594699.0, "step": 14043 }, { "epoch": 1.786541152525124, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.005179405212402, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8738211989402771, "num_tokens": 535627019.0, "step": 14044 }, { "epoch": 1.7866683628037145, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.943299293518066, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8790507316589355, "num_tokens": 535667479.0, "step": 14045 }, { "epoch": 1.786795573082305, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.969558715820312, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8634438514709473, "num_tokens": 535705211.0, "step": 14046 }, { "epoch": 1.7869227833608956, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.996199607849121, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8494706153869629, "num_tokens": 535744907.0, "step": 14047 }, { "epoch": 1.787049993639486, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.920452117919922, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8674590587615967, "num_tokens": 535782526.0, "step": 14048 }, { "epoch": 1.7871772039180764, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.916226387023926, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8848319053649902, "num_tokens": 535824153.0, "step": 14049 }, { "epoch": 1.787304414196667, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.975495338439941, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8812750577926636, "num_tokens": 535859019.0, "step": 14050 }, { "epoch": 1.7874316244752575, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.954034805297852, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.880373477935791, "num_tokens": 535896866.0, "step": 14051 }, { "epoch": 1.787558834753848, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.934267044067383, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8701267838478088, "num_tokens": 535937128.0, "step": 14052 }, { "epoch": 1.7876860450324386, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.984519958496094, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8847523331642151, "num_tokens": 535972727.0, "step": 14053 }, { "epoch": 1.787813255311029, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.915096282958984, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8745005130767822, "num_tokens": 536011491.0, "step": 14054 }, { "epoch": 1.7879404655896196, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.011091232299805, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8724578619003296, "num_tokens": 536048074.0, "step": 14055 }, { "epoch": 1.7880676758682101, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.953718185424805, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8756928443908691, "num_tokens": 536089554.0, "step": 14056 }, { "epoch": 1.7881948861468007, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.988069534301758, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8787208199501038, "num_tokens": 536127672.0, "step": 14057 }, { "epoch": 1.7883220964253912, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.99505615234375, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8700673580169678, "num_tokens": 536157517.0, "step": 14058 }, { "epoch": 1.7884493067039817, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.91507339477539, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8826547265052795, "num_tokens": 536192855.0, "step": 14059 }, { "epoch": 1.7885765169825723, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.997923851013184, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.871228814125061, "num_tokens": 536231727.0, "step": 14060 }, { "epoch": 1.7887037272611628, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.980941772460938, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8826763033866882, "num_tokens": 536270738.0, "step": 14061 }, { "epoch": 1.7888309375397533, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.914402961730957, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8878592848777771, "num_tokens": 536313006.0, "step": 14062 }, { "epoch": 1.7889581478183438, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.974895477294922, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.862973153591156, "num_tokens": 536350697.0, "step": 14063 }, { "epoch": 1.7890853580969344, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.95376205444336, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8778608441352844, "num_tokens": 536394252.0, "step": 14064 }, { "epoch": 1.789212568375525, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.977533340454102, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8684229850769043, "num_tokens": 536426591.0, "step": 14065 }, { "epoch": 1.7893397786541152, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.980829238891602, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8575813174247742, "num_tokens": 536468567.0, "step": 14066 }, { "epoch": 1.7894669889327057, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.946660995483398, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8612192273139954, "num_tokens": 536506352.0, "step": 14067 }, { "epoch": 1.7895941992112963, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.931852340698242, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.863349974155426, "num_tokens": 536538122.0, "step": 14068 }, { "epoch": 1.7897214094898868, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.960493087768555, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8601723313331604, "num_tokens": 536574617.0, "step": 14069 }, { "epoch": 1.7898486197684773, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.88460636138916, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8813517093658447, "num_tokens": 536611895.0, "step": 14070 }, { "epoch": 1.7899758300470678, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.019094467163086, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8629697561264038, "num_tokens": 536651336.0, "step": 14071 }, { "epoch": 1.7901030403256584, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.883667945861816, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8662017583847046, "num_tokens": 536692566.0, "step": 14072 }, { "epoch": 1.7902302506042487, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.928657531738281, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8755264282226562, "num_tokens": 536730855.0, "step": 14073 }, { "epoch": 1.7903574608828392, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.939045906066895, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8683384656906128, "num_tokens": 536766478.0, "step": 14074 }, { "epoch": 1.7904846711614297, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.937394142150879, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8641883134841919, "num_tokens": 536799900.0, "step": 14075 }, { "epoch": 1.7906118814400203, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.932977676391602, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8769869208335876, "num_tokens": 536843597.0, "step": 14076 }, { "epoch": 1.7907390917186108, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.973020553588867, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8811721205711365, "num_tokens": 536879189.0, "step": 14077 }, { "epoch": 1.7908663019972013, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.913346290588379, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8744533658027649, "num_tokens": 536918099.0, "step": 14078 }, { "epoch": 1.7909935122757918, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.941067695617676, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8605672121047974, "num_tokens": 536954410.0, "step": 14079 }, { "epoch": 1.7911207225543824, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.959020614624023, "learning_rate": 1e-06, "loss": 0.3407, "mean_token_accuracy": 0.8919650912284851, "num_tokens": 536986688.0, "step": 14080 }, { "epoch": 1.791247932832973, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.92603874206543, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8696185946464539, "num_tokens": 537024181.0, "step": 14081 }, { "epoch": 1.7913751431115634, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.962357521057129, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8737186789512634, "num_tokens": 537064174.0, "step": 14082 }, { "epoch": 1.791502353390154, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.909337997436523, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8634717464447021, "num_tokens": 537100279.0, "step": 14083 }, { "epoch": 1.7916295636687445, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.000529289245605, "learning_rate": 1e-06, "loss": 0.4721, "mean_token_accuracy": 0.8454412221908569, "num_tokens": 537135339.0, "step": 14084 }, { "epoch": 1.791756773947335, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.933249473571777, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8644859790802002, "num_tokens": 537169608.0, "step": 14085 }, { "epoch": 1.7918839842259255, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.915926933288574, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8781189918518066, "num_tokens": 537205560.0, "step": 14086 }, { "epoch": 1.792011194504516, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.918930053710938, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8587099313735962, "num_tokens": 537238697.0, "step": 14087 }, { "epoch": 1.7921384047831066, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.90214729309082, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8652227520942688, "num_tokens": 537282365.0, "step": 14088 }, { "epoch": 1.7922656150616971, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.947794914245605, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8789418935775757, "num_tokens": 537320805.0, "step": 14089 }, { "epoch": 1.7923928253402877, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.948718070983887, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.872979462146759, "num_tokens": 537354607.0, "step": 14090 }, { "epoch": 1.792520035618878, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.92221736907959, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.879607081413269, "num_tokens": 537386672.0, "step": 14091 }, { "epoch": 1.7926472458974685, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.881916999816895, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.872674822807312, "num_tokens": 537430882.0, "step": 14092 }, { "epoch": 1.792774456176059, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.964194297790527, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8722995519638062, "num_tokens": 537465114.0, "step": 14093 }, { "epoch": 1.7929016664546495, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.924665451049805, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8643958568572998, "num_tokens": 537502820.0, "step": 14094 }, { "epoch": 1.79302887673324, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.99342155456543, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8789182305335999, "num_tokens": 537537422.0, "step": 14095 }, { "epoch": 1.7931560870118306, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.000487327575684, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8751360177993774, "num_tokens": 537566828.0, "step": 14096 }, { "epoch": 1.793283297290421, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.910959243774414, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8674020767211914, "num_tokens": 537606102.0, "step": 14097 }, { "epoch": 1.7934105075690114, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.986322402954102, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8767687678337097, "num_tokens": 537638527.0, "step": 14098 }, { "epoch": 1.793537717847602, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.929778099060059, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8693863153457642, "num_tokens": 537683375.0, "step": 14099 }, { "epoch": 1.7936649281261925, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.910700798034668, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8648504018783569, "num_tokens": 537731354.0, "step": 14100 }, { "epoch": 1.793792138404783, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.952266693115234, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8653415441513062, "num_tokens": 537770055.0, "step": 14101 }, { "epoch": 1.7939193486833735, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.02174186706543, "learning_rate": 1e-06, "loss": 0.4722, "mean_token_accuracy": 0.8504130244255066, "num_tokens": 537807977.0, "step": 14102 }, { "epoch": 1.794046558961964, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.99814224243164, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8650886416435242, "num_tokens": 537847809.0, "step": 14103 }, { "epoch": 1.7941737692405546, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.947367668151855, "learning_rate": 1e-06, "loss": 0.4694, "mean_token_accuracy": 0.8511036038398743, "num_tokens": 537885506.0, "step": 14104 }, { "epoch": 1.7943009795191451, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.029658317565918, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8659943342208862, "num_tokens": 537918936.0, "step": 14105 }, { "epoch": 1.7944281897977357, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.00739574432373, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.874218225479126, "num_tokens": 537954205.0, "step": 14106 }, { "epoch": 1.7945554000763262, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.943395614624023, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8719571828842163, "num_tokens": 537989190.0, "step": 14107 }, { "epoch": 1.7946826103549167, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.969742774963379, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8604000210762024, "num_tokens": 538028852.0, "step": 14108 }, { "epoch": 1.7948098206335072, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.933789253234863, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8756309151649475, "num_tokens": 538061245.0, "step": 14109 }, { "epoch": 1.7949370309120978, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.923055648803711, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8819687366485596, "num_tokens": 538103238.0, "step": 14110 }, { "epoch": 1.7950642411906883, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.95716381072998, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.855649471282959, "num_tokens": 538142692.0, "step": 14111 }, { "epoch": 1.7951914514692788, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.972399711608887, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.867308497428894, "num_tokens": 538183155.0, "step": 14112 }, { "epoch": 1.7953186617478694, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.97020435333252, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8687001466751099, "num_tokens": 538222063.0, "step": 14113 }, { "epoch": 1.7954458720264599, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.964978218078613, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8714582324028015, "num_tokens": 538259847.0, "step": 14114 }, { "epoch": 1.7955730823050502, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.972899436950684, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8767244815826416, "num_tokens": 538298557.0, "step": 14115 }, { "epoch": 1.7957002925836407, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.959320068359375, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8670248985290527, "num_tokens": 538337926.0, "step": 14116 }, { "epoch": 1.7958275028622313, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.949100494384766, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8739908933639526, "num_tokens": 538382755.0, "step": 14117 }, { "epoch": 1.7959547131408218, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.99809455871582, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.851800262928009, "num_tokens": 538417639.0, "step": 14118 }, { "epoch": 1.7960819234194123, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.96230697631836, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.863640546798706, "num_tokens": 538458444.0, "step": 14119 }, { "epoch": 1.7962091336980028, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.921314239501953, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8702806234359741, "num_tokens": 538496580.0, "step": 14120 }, { "epoch": 1.7963363439765934, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.913468360900879, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.878000020980835, "num_tokens": 538538474.0, "step": 14121 }, { "epoch": 1.7964635542551837, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.975337028503418, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8722104430198669, "num_tokens": 538576828.0, "step": 14122 }, { "epoch": 1.7965907645337742, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.998223304748535, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8777444362640381, "num_tokens": 538611630.0, "step": 14123 }, { "epoch": 1.7967179748123647, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.94330883026123, "learning_rate": 1e-06, "loss": 0.4952, "mean_token_accuracy": 0.8459430932998657, "num_tokens": 538659537.0, "step": 14124 }, { "epoch": 1.7968451850909553, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.95451545715332, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8771202564239502, "num_tokens": 538700216.0, "step": 14125 }, { "epoch": 1.7969723953695458, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.987937927246094, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8665218353271484, "num_tokens": 538737719.0, "step": 14126 }, { "epoch": 1.7970996056481363, "ewc_loss": 0.031494140625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 12.92597770690918, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8780442476272583, "num_tokens": 538774723.0, "step": 14127 }, { "epoch": 1.7972268159267268, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.951143264770508, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8795650601387024, "num_tokens": 538815288.0, "step": 14128 }, { "epoch": 1.7973540262053174, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.940545082092285, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.868746280670166, "num_tokens": 538858239.0, "step": 14129 }, { "epoch": 1.797481236483908, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.945683479309082, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8732378482818604, "num_tokens": 538896947.0, "step": 14130 }, { "epoch": 1.7976084467624984, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.020943641662598, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8671269416809082, "num_tokens": 538936511.0, "step": 14131 }, { "epoch": 1.797735657041089, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.953487396240234, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8635321855545044, "num_tokens": 538969198.0, "step": 14132 }, { "epoch": 1.7978628673196795, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.978601455688477, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8810678124427795, "num_tokens": 539002258.0, "step": 14133 }, { "epoch": 1.79799007759827, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.972116470336914, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8565226197242737, "num_tokens": 539039993.0, "step": 14134 }, { "epoch": 1.7981172878768605, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.004525184631348, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8662098050117493, "num_tokens": 539080713.0, "step": 14135 }, { "epoch": 1.798244498155451, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.929490089416504, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8803041577339172, "num_tokens": 539121115.0, "step": 14136 }, { "epoch": 1.7983717084340416, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.028854370117188, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8882222175598145, "num_tokens": 539159889.0, "step": 14137 }, { "epoch": 1.7984989187126321, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.93314266204834, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.868884801864624, "num_tokens": 539201508.0, "step": 14138 }, { "epoch": 1.7986261289912227, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.970104217529297, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.852073073387146, "num_tokens": 539238640.0, "step": 14139 }, { "epoch": 1.798753339269813, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.945892333984375, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8655974864959717, "num_tokens": 539274797.0, "step": 14140 }, { "epoch": 1.7988805495484035, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.017163276672363, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8590694665908813, "num_tokens": 539308410.0, "step": 14141 }, { "epoch": 1.799007759826994, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.958605766296387, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8613645434379578, "num_tokens": 539345578.0, "step": 14142 }, { "epoch": 1.7991349701055845, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.058918952941895, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8545024394989014, "num_tokens": 539378946.0, "step": 14143 }, { "epoch": 1.799262180384175, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.926629066467285, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8814384937286377, "num_tokens": 539417663.0, "step": 14144 }, { "epoch": 1.7993893906627656, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.978649139404297, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8846051096916199, "num_tokens": 539459597.0, "step": 14145 }, { "epoch": 1.799516600941356, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.968853950500488, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8722293972969055, "num_tokens": 539499579.0, "step": 14146 }, { "epoch": 1.7996438112199464, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.923253059387207, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8795562982559204, "num_tokens": 539538505.0, "step": 14147 }, { "epoch": 1.799771021498537, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.009623527526855, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8637158870697021, "num_tokens": 539574824.0, "step": 14148 }, { "epoch": 1.7998982317771275, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.93442153930664, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8568490743637085, "num_tokens": 539614181.0, "step": 14149 }, { "epoch": 1.800025442055718, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.93985652923584, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.857410192489624, "num_tokens": 539656713.0, "step": 14150 }, { "epoch": 1.8001526523343085, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.017142295837402, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8682394027709961, "num_tokens": 539695436.0, "step": 14151 }, { "epoch": 1.800279862612899, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.927348136901855, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8715082406997681, "num_tokens": 539736549.0, "step": 14152 }, { "epoch": 1.8004070728914896, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.93635082244873, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8752216100692749, "num_tokens": 539776326.0, "step": 14153 }, { "epoch": 1.8005342831700801, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.940912246704102, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8761324882507324, "num_tokens": 539819665.0, "step": 14154 }, { "epoch": 1.8006614934486707, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.982200622558594, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8681329488754272, "num_tokens": 539858175.0, "step": 14155 }, { "epoch": 1.8007887037272612, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.95141887664795, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8685956001281738, "num_tokens": 539895152.0, "step": 14156 }, { "epoch": 1.8009159140058517, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.009326934814453, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8745001554489136, "num_tokens": 539941033.0, "step": 14157 }, { "epoch": 1.8010431242844422, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.941472053527832, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8719890713691711, "num_tokens": 539977959.0, "step": 14158 }, { "epoch": 1.8011703345630328, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.03798770904541, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8678485155105591, "num_tokens": 540014857.0, "step": 14159 }, { "epoch": 1.8012975448416233, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.91458797454834, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8685107231140137, "num_tokens": 540055775.0, "step": 14160 }, { "epoch": 1.8014247551202138, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.9847412109375, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.87933349609375, "num_tokens": 540090287.0, "step": 14161 }, { "epoch": 1.8015519653988044, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.957980155944824, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8775553703308105, "num_tokens": 540129139.0, "step": 14162 }, { "epoch": 1.8016791756773949, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.97413158416748, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.869165301322937, "num_tokens": 540165187.0, "step": 14163 }, { "epoch": 1.8018063859559852, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.972474098205566, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8825075030326843, "num_tokens": 540201113.0, "step": 14164 }, { "epoch": 1.8019335962345757, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.992682456970215, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8719416856765747, "num_tokens": 540237448.0, "step": 14165 }, { "epoch": 1.8020608065131662, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.984457969665527, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8631381988525391, "num_tokens": 540279703.0, "step": 14166 }, { "epoch": 1.8021880167917568, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.00347900390625, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8520816564559937, "num_tokens": 540313071.0, "step": 14167 }, { "epoch": 1.8023152270703473, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.937590599060059, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8741976022720337, "num_tokens": 540353940.0, "step": 14168 }, { "epoch": 1.8024424373489378, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.978816032409668, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8672545552253723, "num_tokens": 540391241.0, "step": 14169 }, { "epoch": 1.8025696476275284, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.965553283691406, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8807989358901978, "num_tokens": 540420205.0, "step": 14170 }, { "epoch": 1.8026968579061187, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.937945365905762, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8733407258987427, "num_tokens": 540452121.0, "step": 14171 }, { "epoch": 1.8028240681847092, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.952703475952148, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8697060346603394, "num_tokens": 540487415.0, "step": 14172 }, { "epoch": 1.8029512784632997, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.951677322387695, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8587936162948608, "num_tokens": 540522293.0, "step": 14173 }, { "epoch": 1.8030784887418903, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.95856761932373, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8814797401428223, "num_tokens": 540560345.0, "step": 14174 }, { "epoch": 1.8032056990204808, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.922701835632324, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8780627250671387, "num_tokens": 540598299.0, "step": 14175 }, { "epoch": 1.8033329092990713, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.979869842529297, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8814558386802673, "num_tokens": 540634480.0, "step": 14176 }, { "epoch": 1.8034601195776618, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.946175575256348, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8613604307174683, "num_tokens": 540679473.0, "step": 14177 }, { "epoch": 1.8035873298562524, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.919893264770508, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8629622459411621, "num_tokens": 540722259.0, "step": 14178 }, { "epoch": 1.803714540134843, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.93811321258545, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8733242750167847, "num_tokens": 540758058.0, "step": 14179 }, { "epoch": 1.8038417504134334, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.913559913635254, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8647319078445435, "num_tokens": 540798535.0, "step": 14180 }, { "epoch": 1.803968960692024, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.941776275634766, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8708301782608032, "num_tokens": 540839423.0, "step": 14181 }, { "epoch": 1.8040961709706145, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.946524620056152, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8762614727020264, "num_tokens": 540887512.0, "step": 14182 }, { "epoch": 1.804223381249205, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.952132225036621, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8720096945762634, "num_tokens": 540922292.0, "step": 14183 }, { "epoch": 1.8043505915277955, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 12.991408348083496, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8761004209518433, "num_tokens": 540964105.0, "step": 14184 }, { "epoch": 1.804477801806386, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.999808311462402, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8689614534378052, "num_tokens": 541001793.0, "step": 14185 }, { "epoch": 1.8046050120849766, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.004401206970215, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8770092129707336, "num_tokens": 541041104.0, "step": 14186 }, { "epoch": 1.8047322223635671, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.98088264465332, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8640167713165283, "num_tokens": 541079643.0, "step": 14187 }, { "epoch": 1.8048594326421576, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.970010757446289, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8752669095993042, "num_tokens": 541118554.0, "step": 14188 }, { "epoch": 1.804986642920748, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.959112167358398, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8588128089904785, "num_tokens": 541156893.0, "step": 14189 }, { "epoch": 1.8051138531993385, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.992229461669922, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.88529372215271, "num_tokens": 541197283.0, "step": 14190 }, { "epoch": 1.805241063477929, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.970462799072266, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8834050893783569, "num_tokens": 541232343.0, "step": 14191 }, { "epoch": 1.8053682737565195, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.022156715393066, "learning_rate": 1e-06, "loss": 0.4437, "mean_token_accuracy": 0.8617322444915771, "num_tokens": 541270274.0, "step": 14192 }, { "epoch": 1.80549548403511, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.99463939666748, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.865023672580719, "num_tokens": 541306018.0, "step": 14193 }, { "epoch": 1.8056226943137006, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.966385841369629, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8695365190505981, "num_tokens": 541344974.0, "step": 14194 }, { "epoch": 1.805749904592291, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.01615047454834, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8797298073768616, "num_tokens": 541378590.0, "step": 14195 }, { "epoch": 1.8058771148708814, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.929496765136719, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8718916177749634, "num_tokens": 541415945.0, "step": 14196 }, { "epoch": 1.806004325149472, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.988768577575684, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8671483993530273, "num_tokens": 541459577.0, "step": 14197 }, { "epoch": 1.8061315354280625, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.995738983154297, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8713828325271606, "num_tokens": 541501041.0, "step": 14198 }, { "epoch": 1.806258745706653, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.984979629516602, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8631843328475952, "num_tokens": 541538587.0, "step": 14199 }, { "epoch": 1.8063859559852435, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.040914535522461, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8578871488571167, "num_tokens": 541582225.0, "step": 14200 }, { "epoch": 1.806513166263834, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.003786087036133, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8797224164009094, "num_tokens": 541616956.0, "step": 14201 }, { "epoch": 1.8066403765424246, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.008660316467285, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8587850332260132, "num_tokens": 541650814.0, "step": 14202 }, { "epoch": 1.8067675868210151, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.00502872467041, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8791812658309937, "num_tokens": 541692167.0, "step": 14203 }, { "epoch": 1.8068947970996057, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.991214752197266, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8611339330673218, "num_tokens": 541732427.0, "step": 14204 }, { "epoch": 1.8070220073781962, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.034177780151367, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8706057071685791, "num_tokens": 541773256.0, "step": 14205 }, { "epoch": 1.8071492176567867, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.971744537353516, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8758654594421387, "num_tokens": 541815008.0, "step": 14206 }, { "epoch": 1.8072764279353772, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.98432445526123, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8731951117515564, "num_tokens": 541853462.0, "step": 14207 }, { "epoch": 1.8074036382139678, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.049344062805176, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8651167750358582, "num_tokens": 541890487.0, "step": 14208 }, { "epoch": 1.8075308484925583, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.025167465209961, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8742356300354004, "num_tokens": 541929857.0, "step": 14209 }, { "epoch": 1.8076580587711488, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.993183135986328, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8636751174926758, "num_tokens": 541969923.0, "step": 14210 }, { "epoch": 1.8077852690497394, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.027543067932129, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8786548376083374, "num_tokens": 542008802.0, "step": 14211 }, { "epoch": 1.8079124793283299, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.968048095703125, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8836508393287659, "num_tokens": 542049276.0, "step": 14212 }, { "epoch": 1.8080396896069202, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.998323440551758, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8631523847579956, "num_tokens": 542086921.0, "step": 14213 }, { "epoch": 1.8081668998855107, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.996428489685059, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8702424764633179, "num_tokens": 542124291.0, "step": 14214 }, { "epoch": 1.8082941101641012, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.980993270874023, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8592065572738647, "num_tokens": 542167600.0, "step": 14215 }, { "epoch": 1.8084213204426918, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.02864933013916, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8655065298080444, "num_tokens": 542201134.0, "step": 14216 }, { "epoch": 1.8085485307212823, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.989547729492188, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8567284941673279, "num_tokens": 542242590.0, "step": 14217 }, { "epoch": 1.8086757409998728, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.971744537353516, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8693211078643799, "num_tokens": 542281945.0, "step": 14218 }, { "epoch": 1.8088029512784631, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.979722023010254, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8779551982879639, "num_tokens": 542319389.0, "step": 14219 }, { "epoch": 1.8089301615570537, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.00949478149414, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.857519268989563, "num_tokens": 542356912.0, "step": 14220 }, { "epoch": 1.8090573718356442, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.99496078491211, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8635265827178955, "num_tokens": 542394054.0, "step": 14221 }, { "epoch": 1.8091845821142347, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.055702209472656, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8751381635665894, "num_tokens": 542431063.0, "step": 14222 }, { "epoch": 1.8093117923928252, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.963630676269531, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8592351675033569, "num_tokens": 542471398.0, "step": 14223 }, { "epoch": 1.8094390026714158, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.015414237976074, "learning_rate": 1e-06, "loss": 0.4613, "mean_token_accuracy": 0.8554644584655762, "num_tokens": 542516288.0, "step": 14224 }, { "epoch": 1.8095662129500063, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.979286193847656, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8686609268188477, "num_tokens": 542552292.0, "step": 14225 }, { "epoch": 1.8096934232285968, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.975042343139648, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8723938465118408, "num_tokens": 542592380.0, "step": 14226 }, { "epoch": 1.8098206335071874, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.991329193115234, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8627235293388367, "num_tokens": 542632698.0, "step": 14227 }, { "epoch": 1.8099478437857779, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.981144905090332, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8581509590148926, "num_tokens": 542668234.0, "step": 14228 }, { "epoch": 1.8100750540643684, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.993244171142578, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8679704666137695, "num_tokens": 542706205.0, "step": 14229 }, { "epoch": 1.810202264342959, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.012215614318848, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8677736520767212, "num_tokens": 542737068.0, "step": 14230 }, { "epoch": 1.8103294746215495, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.994556427001953, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8809241056442261, "num_tokens": 542769776.0, "step": 14231 }, { "epoch": 1.81045668490014, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.014108657836914, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8823282122612, "num_tokens": 542802440.0, "step": 14232 }, { "epoch": 1.8105838951787305, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.970684051513672, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8817826509475708, "num_tokens": 542841158.0, "step": 14233 }, { "epoch": 1.810711105457321, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.012617111206055, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.865459680557251, "num_tokens": 542876637.0, "step": 14234 }, { "epoch": 1.8108383157359116, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.970256805419922, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.877364993095398, "num_tokens": 542915087.0, "step": 14235 }, { "epoch": 1.8109655260145021, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.993804931640625, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8616697192192078, "num_tokens": 542950635.0, "step": 14236 }, { "epoch": 1.8110927362930926, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.06216049194336, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8654199242591858, "num_tokens": 542990078.0, "step": 14237 }, { "epoch": 1.811219946571683, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.978686332702637, "learning_rate": 1e-06, "loss": 0.4813, "mean_token_accuracy": 0.8452487587928772, "num_tokens": 543030333.0, "step": 14238 }, { "epoch": 1.8113471568502735, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.99744701385498, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8624134063720703, "num_tokens": 543070537.0, "step": 14239 }, { "epoch": 1.811474367128864, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.03408432006836, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8678953647613525, "num_tokens": 543109137.0, "step": 14240 }, { "epoch": 1.8116015774074545, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.965571403503418, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8627278804779053, "num_tokens": 543146616.0, "step": 14241 }, { "epoch": 1.811728787686045, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.038125991821289, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8830043077468872, "num_tokens": 543178433.0, "step": 14242 }, { "epoch": 1.8118559979646356, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.019194602966309, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8820844888687134, "num_tokens": 543215086.0, "step": 14243 }, { "epoch": 1.811983208243226, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.964365005493164, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8541901111602783, "num_tokens": 543252111.0, "step": 14244 }, { "epoch": 1.8121104185218164, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.014501571655273, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8770124912261963, "num_tokens": 543291938.0, "step": 14245 }, { "epoch": 1.812237628800407, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.961188316345215, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.873769998550415, "num_tokens": 543335141.0, "step": 14246 }, { "epoch": 1.8123648390789975, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.04223346710205, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8694279789924622, "num_tokens": 543371305.0, "step": 14247 }, { "epoch": 1.812492049357588, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.00258731842041, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8797803521156311, "num_tokens": 543407431.0, "step": 14248 }, { "epoch": 1.8126192596361785, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.983407020568848, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8696091175079346, "num_tokens": 543450287.0, "step": 14249 }, { "epoch": 1.812746469914769, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.999483108520508, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8823496103286743, "num_tokens": 543486521.0, "step": 14250 }, { "epoch": 1.8128736801933596, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.988677024841309, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8691484928131104, "num_tokens": 543522157.0, "step": 14251 }, { "epoch": 1.8130008904719501, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.007068634033203, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.872200608253479, "num_tokens": 543560503.0, "step": 14252 }, { "epoch": 1.8131281007505406, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.064070701599121, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.857651948928833, "num_tokens": 543598943.0, "step": 14253 }, { "epoch": 1.8132553110291312, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.02733039855957, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8716577291488647, "num_tokens": 543640466.0, "step": 14254 }, { "epoch": 1.8133825213077217, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.953454971313477, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8712283372879028, "num_tokens": 543677581.0, "step": 14255 }, { "epoch": 1.8135097315863122, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.055607795715332, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8620136976242065, "num_tokens": 543711220.0, "step": 14256 }, { "epoch": 1.8136369418649028, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.979702949523926, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8636873960494995, "num_tokens": 543750109.0, "step": 14257 }, { "epoch": 1.8137641521434933, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.982304573059082, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8747243881225586, "num_tokens": 543788040.0, "step": 14258 }, { "epoch": 1.8138913624220838, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.01406192779541, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8694688081741333, "num_tokens": 543821037.0, "step": 14259 }, { "epoch": 1.8140185727006743, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.974381446838379, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8835971355438232, "num_tokens": 543856331.0, "step": 14260 }, { "epoch": 1.8141457829792649, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.978797912597656, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8735823631286621, "num_tokens": 543892898.0, "step": 14261 }, { "epoch": 1.8142729932578552, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.972851753234863, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8662235140800476, "num_tokens": 543932858.0, "step": 14262 }, { "epoch": 1.8144002035364457, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.97872257232666, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8787845373153687, "num_tokens": 543968145.0, "step": 14263 }, { "epoch": 1.8145274138150362, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.005617141723633, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8684378862380981, "num_tokens": 544000495.0, "step": 14264 }, { "epoch": 1.8146546240936268, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.960907936096191, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.865289568901062, "num_tokens": 544041345.0, "step": 14265 }, { "epoch": 1.8147818343722173, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.044205665588379, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8741334080696106, "num_tokens": 544080084.0, "step": 14266 }, { "epoch": 1.8149090446508078, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.935614585876465, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8618254065513611, "num_tokens": 544115828.0, "step": 14267 }, { "epoch": 1.8150362549293981, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.997933387756348, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8660671710968018, "num_tokens": 544153213.0, "step": 14268 }, { "epoch": 1.8151634652079887, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.973980903625488, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8731933236122131, "num_tokens": 544192383.0, "step": 14269 }, { "epoch": 1.8152906754865792, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.005106925964355, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8714653253555298, "num_tokens": 544232034.0, "step": 14270 }, { "epoch": 1.8154178857651697, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.941425323486328, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8536927700042725, "num_tokens": 544266242.0, "step": 14271 }, { "epoch": 1.8155450960437602, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.996781349182129, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8595677614212036, "num_tokens": 544304995.0, "step": 14272 }, { "epoch": 1.8156723063223508, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.046184539794922, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8691417574882507, "num_tokens": 544341950.0, "step": 14273 }, { "epoch": 1.8157995166009413, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.926435470581055, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8650878071784973, "num_tokens": 544379535.0, "step": 14274 }, { "epoch": 1.8159267268795318, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.970717430114746, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8680694699287415, "num_tokens": 544416939.0, "step": 14275 }, { "epoch": 1.8160539371581224, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.997671127319336, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8805361390113831, "num_tokens": 544461476.0, "step": 14276 }, { "epoch": 1.8161811474367129, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.966376304626465, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8696243762969971, "num_tokens": 544505051.0, "step": 14277 }, { "epoch": 1.8163083577153034, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.99819564819336, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8698415756225586, "num_tokens": 544550408.0, "step": 14278 }, { "epoch": 1.816435567993894, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.98609733581543, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8693481087684631, "num_tokens": 544586645.0, "step": 14279 }, { "epoch": 1.8165627782724845, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.026862144470215, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8628582954406738, "num_tokens": 544628087.0, "step": 14280 }, { "epoch": 1.816689988551075, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.98798656463623, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8845345973968506, "num_tokens": 544667961.0, "step": 14281 }, { "epoch": 1.8168171988296655, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.989381790161133, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8668744564056396, "num_tokens": 544704919.0, "step": 14282 }, { "epoch": 1.816944409108256, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.016607284545898, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8744341135025024, "num_tokens": 544741421.0, "step": 14283 }, { "epoch": 1.8170716193868466, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.022172927856445, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8740736842155457, "num_tokens": 544776414.0, "step": 14284 }, { "epoch": 1.817198829665437, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.918009757995605, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8620142936706543, "num_tokens": 544814964.0, "step": 14285 }, { "epoch": 1.8173260399440276, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.036065101623535, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8555158376693726, "num_tokens": 544853381.0, "step": 14286 }, { "epoch": 1.817453250222618, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.996428489685059, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8724659085273743, "num_tokens": 544892474.0, "step": 14287 }, { "epoch": 1.8175804605012085, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.006731033325195, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8767954111099243, "num_tokens": 544927667.0, "step": 14288 }, { "epoch": 1.817707670779799, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.044316291809082, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8754738569259644, "num_tokens": 544959780.0, "step": 14289 }, { "epoch": 1.8178348810583895, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.006622314453125, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8626166582107544, "num_tokens": 545001832.0, "step": 14290 }, { "epoch": 1.81796209133698, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.93978214263916, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8779767155647278, "num_tokens": 545038489.0, "step": 14291 }, { "epoch": 1.8180893016155706, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.11654281616211, "learning_rate": 1e-06, "loss": 0.4931, "mean_token_accuracy": 0.8496595621109009, "num_tokens": 545069086.0, "step": 14292 }, { "epoch": 1.818216511894161, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.017044067382812, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8721451759338379, "num_tokens": 545104564.0, "step": 14293 }, { "epoch": 1.8183437221727514, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.973113059997559, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.873534083366394, "num_tokens": 545149036.0, "step": 14294 }, { "epoch": 1.818470932451342, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.014348030090332, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.881244957447052, "num_tokens": 545183825.0, "step": 14295 }, { "epoch": 1.8185981427299325, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.97177791595459, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8729990720748901, "num_tokens": 545221274.0, "step": 14296 }, { "epoch": 1.818725353008523, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.98934555053711, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8767127990722656, "num_tokens": 545259242.0, "step": 14297 }, { "epoch": 1.8188525632871135, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.044618606567383, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8570463061332703, "num_tokens": 545289380.0, "step": 14298 }, { "epoch": 1.818979773565704, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.976168632507324, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8734169006347656, "num_tokens": 545321935.0, "step": 14299 }, { "epoch": 1.8191069838442946, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.998650550842285, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8772081136703491, "num_tokens": 545365714.0, "step": 14300 }, { "epoch": 1.8192341941228851, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.973506927490234, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8595156073570251, "num_tokens": 545406838.0, "step": 14301 }, { "epoch": 1.8193614044014756, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.051556587219238, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8672632575035095, "num_tokens": 545443265.0, "step": 14302 }, { "epoch": 1.8194886146800662, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.981616973876953, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8805111050605774, "num_tokens": 545479903.0, "step": 14303 }, { "epoch": 1.8196158249586567, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.040403366088867, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8762797713279724, "num_tokens": 545513767.0, "step": 14304 }, { "epoch": 1.8197430352372472, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.053088188171387, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.854952335357666, "num_tokens": 545550541.0, "step": 14305 }, { "epoch": 1.8198702455158378, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.927631378173828, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8711185455322266, "num_tokens": 545590739.0, "step": 14306 }, { "epoch": 1.8199974557944283, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.04904556274414, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8562431335449219, "num_tokens": 545627127.0, "step": 14307 }, { "epoch": 1.8201246660730188, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.066898345947266, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8745386600494385, "num_tokens": 545665473.0, "step": 14308 }, { "epoch": 1.8202518763516093, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.9381742477417, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8852759003639221, "num_tokens": 545708018.0, "step": 14309 }, { "epoch": 1.8203790866301999, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.067914009094238, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8752526640892029, "num_tokens": 545745635.0, "step": 14310 }, { "epoch": 1.8205062969087902, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.991678237915039, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8587393760681152, "num_tokens": 545785260.0, "step": 14311 }, { "epoch": 1.8206335071873807, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.986835479736328, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.867145836353302, "num_tokens": 545821696.0, "step": 14312 }, { "epoch": 1.8207607174659712, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.076961517333984, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.883173942565918, "num_tokens": 545857971.0, "step": 14313 }, { "epoch": 1.8208879277445618, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.007043838500977, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8595361113548279, "num_tokens": 545891000.0, "step": 14314 }, { "epoch": 1.8210151380231523, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.030783653259277, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8567090630531311, "num_tokens": 545932825.0, "step": 14315 }, { "epoch": 1.8211423483017428, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.000112533569336, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8596838712692261, "num_tokens": 545972849.0, "step": 14316 }, { "epoch": 1.8212695585803331, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.954721450805664, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8822388648986816, "num_tokens": 546011492.0, "step": 14317 }, { "epoch": 1.8213967688589237, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.063924789428711, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8638930916786194, "num_tokens": 546048461.0, "step": 14318 }, { "epoch": 1.8215239791375142, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.924871444702148, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8788244128227234, "num_tokens": 546087113.0, "step": 14319 }, { "epoch": 1.8216511894161047, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.036264419555664, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8725252747535706, "num_tokens": 546124413.0, "step": 14320 }, { "epoch": 1.8217783996946952, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.980622291564941, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8815823793411255, "num_tokens": 546165763.0, "step": 14321 }, { "epoch": 1.8219056099732858, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.923508644104004, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8703978061676025, "num_tokens": 546203218.0, "step": 14322 }, { "epoch": 1.8220328202518763, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.012552261352539, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8602477312088013, "num_tokens": 546247610.0, "step": 14323 }, { "epoch": 1.8221600305304668, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.981999397277832, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8617894053459167, "num_tokens": 546288952.0, "step": 14324 }, { "epoch": 1.8222872408090574, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.0010404586792, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8810926079750061, "num_tokens": 546322449.0, "step": 14325 }, { "epoch": 1.8224144510876479, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.905794143676758, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8760544061660767, "num_tokens": 546362692.0, "step": 14326 }, { "epoch": 1.8225416613662384, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.040241241455078, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8535911440849304, "num_tokens": 546398856.0, "step": 14327 }, { "epoch": 1.822668871644829, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.930916786193848, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8621571063995361, "num_tokens": 546439633.0, "step": 14328 }, { "epoch": 1.8227960819234195, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.028873443603516, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.874235987663269, "num_tokens": 546477282.0, "step": 14329 }, { "epoch": 1.82292329220201, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.980836868286133, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8821573257446289, "num_tokens": 546514584.0, "step": 14330 }, { "epoch": 1.8230505024806005, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.982251167297363, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8739851713180542, "num_tokens": 546550761.0, "step": 14331 }, { "epoch": 1.823177712759191, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.033642768859863, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8561846613883972, "num_tokens": 546594157.0, "step": 14332 }, { "epoch": 1.8233049230377816, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.987070083618164, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8776207566261292, "num_tokens": 546634671.0, "step": 14333 }, { "epoch": 1.823432133316372, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.048043251037598, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.870476484298706, "num_tokens": 546672084.0, "step": 14334 }, { "epoch": 1.8235593435949626, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.025238990783691, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8640053272247314, "num_tokens": 546701920.0, "step": 14335 }, { "epoch": 1.823686553873553, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.998922348022461, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8747173547744751, "num_tokens": 546735776.0, "step": 14336 }, { "epoch": 1.8238137641521435, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.016273498535156, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8751403093338013, "num_tokens": 546772991.0, "step": 14337 }, { "epoch": 1.823940974430734, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.073554039001465, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.86355060338974, "num_tokens": 546805310.0, "step": 14338 }, { "epoch": 1.8240681847093245, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.001457214355469, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8763678073883057, "num_tokens": 546838240.0, "step": 14339 }, { "epoch": 1.824195394987915, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.988469123840332, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8657352924346924, "num_tokens": 546878484.0, "step": 14340 }, { "epoch": 1.8243226052665056, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.992706298828125, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8850502967834473, "num_tokens": 546919747.0, "step": 14341 }, { "epoch": 1.8244498155450959, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.971122741699219, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8695895671844482, "num_tokens": 546962614.0, "step": 14342 }, { "epoch": 1.8245770258236864, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.028246879577637, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8634077906608582, "num_tokens": 547000059.0, "step": 14343 }, { "epoch": 1.824704236102277, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.958255767822266, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8710037469863892, "num_tokens": 547037372.0, "step": 14344 }, { "epoch": 1.8248314463808675, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.979161262512207, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8570308089256287, "num_tokens": 547071840.0, "step": 14345 }, { "epoch": 1.824958656659458, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.988036155700684, "learning_rate": 1e-06, "loss": 0.4891, "mean_token_accuracy": 0.8450480699539185, "num_tokens": 547108916.0, "step": 14346 }, { "epoch": 1.8250858669380485, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.924921035766602, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8873066902160645, "num_tokens": 547146991.0, "step": 14347 }, { "epoch": 1.825213077216639, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.005401611328125, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8803185224533081, "num_tokens": 547184093.0, "step": 14348 }, { "epoch": 1.8253402874952296, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.992782592773438, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8761032223701477, "num_tokens": 547217570.0, "step": 14349 }, { "epoch": 1.8254674977738201, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.953822135925293, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8691616058349609, "num_tokens": 547259498.0, "step": 14350 }, { "epoch": 1.8255947080524106, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.002567291259766, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8692814111709595, "num_tokens": 547298419.0, "step": 14351 }, { "epoch": 1.8257219183310012, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.93148136138916, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8776167631149292, "num_tokens": 547342984.0, "step": 14352 }, { "epoch": 1.8258491286095917, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.03608226776123, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8712904453277588, "num_tokens": 547383162.0, "step": 14353 }, { "epoch": 1.8259763388881822, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.956315040588379, "learning_rate": 1e-06, "loss": 0.3417, "mean_token_accuracy": 0.8908572196960449, "num_tokens": 547416967.0, "step": 14354 }, { "epoch": 1.8261035491667728, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.051788330078125, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8732803463935852, "num_tokens": 547454695.0, "step": 14355 }, { "epoch": 1.8262307594453633, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.040257453918457, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8803009986877441, "num_tokens": 547494258.0, "step": 14356 }, { "epoch": 1.8263579697239538, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.064733505249023, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.865461528301239, "num_tokens": 547531404.0, "step": 14357 }, { "epoch": 1.8264851800025443, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.026021957397461, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8745303153991699, "num_tokens": 547566553.0, "step": 14358 }, { "epoch": 1.8266123902811349, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.022644996643066, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8632156252861023, "num_tokens": 547605505.0, "step": 14359 }, { "epoch": 1.8267396005597252, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.987154960632324, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8786231875419617, "num_tokens": 547643146.0, "step": 14360 }, { "epoch": 1.8268668108383157, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.003376007080078, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8674651384353638, "num_tokens": 547683373.0, "step": 14361 }, { "epoch": 1.8269940211169062, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.02528190612793, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8656035661697388, "num_tokens": 547716762.0, "step": 14362 }, { "epoch": 1.8271212313954968, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.032000541687012, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8781615495681763, "num_tokens": 547756907.0, "step": 14363 }, { "epoch": 1.8272484416740873, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.962435722351074, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8666276931762695, "num_tokens": 547793545.0, "step": 14364 }, { "epoch": 1.8273756519526778, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.018548965454102, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8610546588897705, "num_tokens": 547836845.0, "step": 14365 }, { "epoch": 1.8275028622312681, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.021263122558594, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8759899139404297, "num_tokens": 547875849.0, "step": 14366 }, { "epoch": 1.8276300725098586, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.998673439025879, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8633552193641663, "num_tokens": 547918672.0, "step": 14367 }, { "epoch": 1.8277572827884492, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.021101951599121, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8577910661697388, "num_tokens": 547958681.0, "step": 14368 }, { "epoch": 1.8278844930670397, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.048935890197754, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8646849989891052, "num_tokens": 547996940.0, "step": 14369 }, { "epoch": 1.8280117033456302, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.010984420776367, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8776861429214478, "num_tokens": 548030188.0, "step": 14370 }, { "epoch": 1.8281389136242208, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.997278213500977, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8856121897697449, "num_tokens": 548066891.0, "step": 14371 }, { "epoch": 1.8282661239028113, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.049758911132812, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8761318325996399, "num_tokens": 548104664.0, "step": 14372 }, { "epoch": 1.8283933341814018, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.990193367004395, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8658578991889954, "num_tokens": 548143938.0, "step": 14373 }, { "epoch": 1.8285205444599923, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.06083869934082, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8684011697769165, "num_tokens": 548183213.0, "step": 14374 }, { "epoch": 1.8286477547385829, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.01951789855957, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.854053258895874, "num_tokens": 548221304.0, "step": 14375 }, { "epoch": 1.8287749650171734, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.039164543151855, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8677045106887817, "num_tokens": 548261123.0, "step": 14376 }, { "epoch": 1.828902175295764, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.006692886352539, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8508962392807007, "num_tokens": 548305040.0, "step": 14377 }, { "epoch": 1.8290293855743545, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.98455810546875, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8774286508560181, "num_tokens": 548338538.0, "step": 14378 }, { "epoch": 1.829156595852945, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.015084266662598, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8729704022407532, "num_tokens": 548378510.0, "step": 14379 }, { "epoch": 1.8292838061315355, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.002145767211914, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8637975454330444, "num_tokens": 548415790.0, "step": 14380 }, { "epoch": 1.829411016410126, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.057024002075195, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8610016703605652, "num_tokens": 548447294.0, "step": 14381 }, { "epoch": 1.8295382266887166, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.004637718200684, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8769291043281555, "num_tokens": 548484617.0, "step": 14382 }, { "epoch": 1.829665436967307, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.014644622802734, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8660879731178284, "num_tokens": 548527552.0, "step": 14383 }, { "epoch": 1.8297926472458976, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.966172218322754, "learning_rate": 1e-06, "loss": 0.3474, "mean_token_accuracy": 0.8887360692024231, "num_tokens": 548561322.0, "step": 14384 }, { "epoch": 1.829919857524488, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.02401065826416, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8804864883422852, "num_tokens": 548594810.0, "step": 14385 }, { "epoch": 1.8300470678030785, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.024746894836426, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8639827370643616, "num_tokens": 548631094.0, "step": 14386 }, { "epoch": 1.830174278081669, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 12.924477577209473, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8714689612388611, "num_tokens": 548669162.0, "step": 14387 }, { "epoch": 1.8303014883602595, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.013346672058105, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8698444366455078, "num_tokens": 548704555.0, "step": 14388 }, { "epoch": 1.83042869863885, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.01504135131836, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8841779828071594, "num_tokens": 548741852.0, "step": 14389 }, { "epoch": 1.8305559089174406, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.989837646484375, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8580661416053772, "num_tokens": 548788455.0, "step": 14390 }, { "epoch": 1.8306831191960309, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.981800079345703, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8773794174194336, "num_tokens": 548826347.0, "step": 14391 }, { "epoch": 1.8308103294746214, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.010102272033691, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8640221953392029, "num_tokens": 548865428.0, "step": 14392 }, { "epoch": 1.830937539753212, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.03488826751709, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8575753569602966, "num_tokens": 548901671.0, "step": 14393 }, { "epoch": 1.8310647500318025, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.040264129638672, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8660805225372314, "num_tokens": 548939234.0, "step": 14394 }, { "epoch": 1.831191960310393, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.005270004272461, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8751276135444641, "num_tokens": 548975334.0, "step": 14395 }, { "epoch": 1.8313191705889835, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.034516334533691, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8635985255241394, "num_tokens": 549016372.0, "step": 14396 }, { "epoch": 1.831446380867574, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.003055572509766, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8770281672477722, "num_tokens": 549058049.0, "step": 14397 }, { "epoch": 1.8315735911461646, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.056803703308105, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8573134541511536, "num_tokens": 549097855.0, "step": 14398 }, { "epoch": 1.831700801424755, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.99128532409668, "learning_rate": 1e-06, "loss": 0.3415, "mean_token_accuracy": 0.8902573585510254, "num_tokens": 549130906.0, "step": 14399 }, { "epoch": 1.8318280117033456, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.061540603637695, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8517419695854187, "num_tokens": 549176574.0, "step": 14400 }, { "epoch": 1.8319552219819362, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.008710861206055, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8687474727630615, "num_tokens": 549213945.0, "step": 14401 }, { "epoch": 1.8320824322605267, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.00790023803711, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8613365888595581, "num_tokens": 549255843.0, "step": 14402 }, { "epoch": 1.8322096425391172, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.022639274597168, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8745124340057373, "num_tokens": 549292573.0, "step": 14403 }, { "epoch": 1.8323368528177078, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.991495132446289, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8824030756950378, "num_tokens": 549335291.0, "step": 14404 }, { "epoch": 1.8324640630962983, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.992105484008789, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8575785756111145, "num_tokens": 549381337.0, "step": 14405 }, { "epoch": 1.8325912733748888, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.010541915893555, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.881207287311554, "num_tokens": 549418180.0, "step": 14406 }, { "epoch": 1.8327184836534793, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.986588478088379, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8780635595321655, "num_tokens": 549461623.0, "step": 14407 }, { "epoch": 1.8328456939320699, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.000646591186523, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8737105131149292, "num_tokens": 549500791.0, "step": 14408 }, { "epoch": 1.8329729042106602, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.0404634475708, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8760279417037964, "num_tokens": 549533343.0, "step": 14409 }, { "epoch": 1.8331001144892507, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.048530578613281, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8619663119316101, "num_tokens": 549572707.0, "step": 14410 }, { "epoch": 1.8332273247678412, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.01106071472168, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.876312255859375, "num_tokens": 549609772.0, "step": 14411 }, { "epoch": 1.8333545350464318, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.05827808380127, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8592982292175293, "num_tokens": 549648626.0, "step": 14412 }, { "epoch": 1.8334817453250223, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.033513069152832, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8583623170852661, "num_tokens": 549693211.0, "step": 14413 }, { "epoch": 1.8336089556036128, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.023972511291504, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8644641637802124, "num_tokens": 549729463.0, "step": 14414 }, { "epoch": 1.8337361658822031, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.035202980041504, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8753935098648071, "num_tokens": 549769139.0, "step": 14415 }, { "epoch": 1.8338633761607936, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.004194259643555, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8740377426147461, "num_tokens": 549807067.0, "step": 14416 }, { "epoch": 1.8339905864393842, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.067451477050781, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8695399761199951, "num_tokens": 549842689.0, "step": 14417 }, { "epoch": 1.8341177967179747, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.0833101272583, "learning_rate": 1e-06, "loss": 0.342, "mean_token_accuracy": 0.889326810836792, "num_tokens": 549876108.0, "step": 14418 }, { "epoch": 1.8342450069965652, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.048243522644043, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8841580152511597, "num_tokens": 549909656.0, "step": 14419 }, { "epoch": 1.8343722172751558, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.062253952026367, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8610345125198364, "num_tokens": 549949156.0, "step": 14420 }, { "epoch": 1.8344994275537463, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.032142639160156, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.876745343208313, "num_tokens": 549990256.0, "step": 14421 }, { "epoch": 1.8346266378323368, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.090456008911133, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8790372610092163, "num_tokens": 550023557.0, "step": 14422 }, { "epoch": 1.8347538481109273, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.0316162109375, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8666379451751709, "num_tokens": 550069402.0, "step": 14423 }, { "epoch": 1.8348810583895179, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.998425483703613, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8776756525039673, "num_tokens": 550110860.0, "step": 14424 }, { "epoch": 1.8350082686681084, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.047612190246582, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8545789122581482, "num_tokens": 550151736.0, "step": 14425 }, { "epoch": 1.835135478946699, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.056145668029785, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8735135197639465, "num_tokens": 550187731.0, "step": 14426 }, { "epoch": 1.8352626892252895, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.071231842041016, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8721569180488586, "num_tokens": 550220545.0, "step": 14427 }, { "epoch": 1.83538989950388, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.044611930847168, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8739407062530518, "num_tokens": 550254755.0, "step": 14428 }, { "epoch": 1.8355171097824705, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.985068321228027, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8771998882293701, "num_tokens": 550295191.0, "step": 14429 }, { "epoch": 1.835644320061061, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.050220489501953, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8648203611373901, "num_tokens": 550332984.0, "step": 14430 }, { "epoch": 1.8357715303396516, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.058297157287598, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8720741868019104, "num_tokens": 550377470.0, "step": 14431 }, { "epoch": 1.835898740618242, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.074848175048828, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8746142387390137, "num_tokens": 550416123.0, "step": 14432 }, { "epoch": 1.8360259508968326, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.114842414855957, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.874788761138916, "num_tokens": 550458062.0, "step": 14433 }, { "epoch": 1.836153161175423, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.116854667663574, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8645485043525696, "num_tokens": 550496485.0, "step": 14434 }, { "epoch": 1.8362803714540135, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.990689277648926, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8723849058151245, "num_tokens": 550536500.0, "step": 14435 }, { "epoch": 1.836407581732604, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.18730640411377, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8651229739189148, "num_tokens": 550575611.0, "step": 14436 }, { "epoch": 1.8365347920111945, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.010180473327637, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8625732660293579, "num_tokens": 550616431.0, "step": 14437 }, { "epoch": 1.836662002289785, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 13.088993072509766, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.880281925201416, "num_tokens": 550653626.0, "step": 14438 }, { "epoch": 1.8367892125683756, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.075533866882324, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8666746616363525, "num_tokens": 550690899.0, "step": 14439 }, { "epoch": 1.8369164228469659, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.02180290222168, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8710119724273682, "num_tokens": 550734408.0, "step": 14440 }, { "epoch": 1.8370436331255564, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.012950897216797, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8579015731811523, "num_tokens": 550767888.0, "step": 14441 }, { "epoch": 1.837170843404147, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.059300422668457, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8721460700035095, "num_tokens": 550807192.0, "step": 14442 }, { "epoch": 1.8372980536827375, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.028307914733887, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8781365752220154, "num_tokens": 550848006.0, "step": 14443 }, { "epoch": 1.837425263961328, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.003787994384766, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8706063628196716, "num_tokens": 550888350.0, "step": 14444 }, { "epoch": 1.8375524742399185, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.088913917541504, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.877111554145813, "num_tokens": 550925389.0, "step": 14445 }, { "epoch": 1.837679684518509, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.049484252929688, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8748830556869507, "num_tokens": 550962610.0, "step": 14446 }, { "epoch": 1.8378068947970996, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.027267456054688, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.866633415222168, "num_tokens": 551001018.0, "step": 14447 }, { "epoch": 1.83793410507569, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.05720043182373, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8720079660415649, "num_tokens": 551038505.0, "step": 14448 }, { "epoch": 1.8380613153542806, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.031627655029297, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8817142248153687, "num_tokens": 551072235.0, "step": 14449 }, { "epoch": 1.8381885256328712, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.014639854431152, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8836676478385925, "num_tokens": 551110129.0, "step": 14450 }, { "epoch": 1.8383157359114617, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.049871444702148, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8713705539703369, "num_tokens": 551152380.0, "step": 14451 }, { "epoch": 1.8384429461900522, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.991690635681152, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8819621205329895, "num_tokens": 551195942.0, "step": 14452 }, { "epoch": 1.8385701564686427, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.047703742980957, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8629190921783447, "num_tokens": 551238502.0, "step": 14453 }, { "epoch": 1.8386973667472333, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.017324447631836, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8803165555000305, "num_tokens": 551278091.0, "step": 14454 }, { "epoch": 1.8388245770258238, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 13.051742553710938, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8508412837982178, "num_tokens": 551314940.0, "step": 14455 }, { "epoch": 1.8389517873044143, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.006153106689453, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8577947020530701, "num_tokens": 551356427.0, "step": 14456 }, { "epoch": 1.8390789975830049, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.996105194091797, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8754212856292725, "num_tokens": 551400436.0, "step": 14457 }, { "epoch": 1.8392062078615952, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.01538372039795, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8696216344833374, "num_tokens": 551438335.0, "step": 14458 }, { "epoch": 1.8393334181401857, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.018607139587402, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8740161657333374, "num_tokens": 551471214.0, "step": 14459 }, { "epoch": 1.8394606284187762, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.049768447875977, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8583587408065796, "num_tokens": 551504071.0, "step": 14460 }, { "epoch": 1.8395878386973668, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.04447078704834, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8659066557884216, "num_tokens": 551538456.0, "step": 14461 }, { "epoch": 1.8397150489759573, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.981164932250977, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.874189019203186, "num_tokens": 551575334.0, "step": 14462 }, { "epoch": 1.8398422592545478, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.06417179107666, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8790323138237, "num_tokens": 551617127.0, "step": 14463 }, { "epoch": 1.8399694695331381, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.071370124816895, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8750163912773132, "num_tokens": 551655548.0, "step": 14464 }, { "epoch": 1.8400966798117286, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.017946243286133, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8677338361740112, "num_tokens": 551691807.0, "step": 14465 }, { "epoch": 1.8402238900903192, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.06328296661377, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8790768980979919, "num_tokens": 551729805.0, "step": 14466 }, { "epoch": 1.8403511003689097, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.04344654083252, "learning_rate": 1e-06, "loss": 0.4874, "mean_token_accuracy": 0.8477950096130371, "num_tokens": 551775220.0, "step": 14467 }, { "epoch": 1.8404783106475002, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.051445007324219, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8623299598693848, "num_tokens": 551816242.0, "step": 14468 }, { "epoch": 1.8406055209260908, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.052207946777344, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8600873947143555, "num_tokens": 551854929.0, "step": 14469 }, { "epoch": 1.8407327312046813, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.994497299194336, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8677579164505005, "num_tokens": 551898877.0, "step": 14470 }, { "epoch": 1.8408599414832718, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.020505905151367, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8871374726295471, "num_tokens": 551934252.0, "step": 14471 }, { "epoch": 1.8409871517618623, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.020275115966797, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8820371627807617, "num_tokens": 551974779.0, "step": 14472 }, { "epoch": 1.8411143620404529, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.952396392822266, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8702455759048462, "num_tokens": 552011627.0, "step": 14473 }, { "epoch": 1.8412415723190434, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.030475616455078, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8497940301895142, "num_tokens": 552045333.0, "step": 14474 }, { "epoch": 1.841368782597634, "ewc_loss": 0.03173828125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 12.856258392333984, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8652746081352234, "num_tokens": 552079974.0, "step": 14475 }, { "epoch": 1.8414959928762245, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.12790298461914, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.878159761428833, "num_tokens": 552117507.0, "step": 14476 }, { "epoch": 1.841623203154815, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 12.933022499084473, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8700643181800842, "num_tokens": 552156009.0, "step": 14477 }, { "epoch": 1.8417504134334055, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.02932357788086, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8791845440864563, "num_tokens": 552189092.0, "step": 14478 }, { "epoch": 1.841877623711996, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 12.947699546813965, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8689660429954529, "num_tokens": 552228305.0, "step": 14479 }, { "epoch": 1.8420048339905866, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.031051635742188, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8721091151237488, "num_tokens": 552269402.0, "step": 14480 }, { "epoch": 1.842132044269177, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.012532234191895, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8681113123893738, "num_tokens": 552304870.0, "step": 14481 }, { "epoch": 1.8422592545477676, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.021695137023926, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8663833737373352, "num_tokens": 552336802.0, "step": 14482 }, { "epoch": 1.842386464826358, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 12.975284576416016, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8837785720825195, "num_tokens": 552376124.0, "step": 14483 }, { "epoch": 1.8425136751049485, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.043937683105469, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8616575598716736, "num_tokens": 552412285.0, "step": 14484 }, { "epoch": 1.842640885383539, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.039661407470703, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8713261485099792, "num_tokens": 552447527.0, "step": 14485 }, { "epoch": 1.8427680956621295, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.009976387023926, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8730946779251099, "num_tokens": 552486569.0, "step": 14486 }, { "epoch": 1.84289530594072, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.006399154663086, "learning_rate": 1e-06, "loss": 0.5416, "mean_token_accuracy": 0.8363186120986938, "num_tokens": 552529158.0, "step": 14487 }, { "epoch": 1.8430225162193106, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 12.995255470275879, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8692381978034973, "num_tokens": 552572183.0, "step": 14488 }, { "epoch": 1.8431497264979009, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 12.987592697143555, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8682796955108643, "num_tokens": 552609785.0, "step": 14489 }, { "epoch": 1.8432769367764914, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.019268989562988, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8767775297164917, "num_tokens": 552641243.0, "step": 14490 }, { "epoch": 1.843404147055082, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.03568172454834, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8621402382850647, "num_tokens": 552677595.0, "step": 14491 }, { "epoch": 1.8435313573336725, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.057360649108887, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8715894222259521, "num_tokens": 552715492.0, "step": 14492 }, { "epoch": 1.843658567612263, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 12.977716445922852, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.869495153427124, "num_tokens": 552757410.0, "step": 14493 }, { "epoch": 1.8437857778908535, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.003400802612305, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8850390911102295, "num_tokens": 552793794.0, "step": 14494 }, { "epoch": 1.843912988169444, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 12.993027687072754, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8554348945617676, "num_tokens": 552836010.0, "step": 14495 }, { "epoch": 1.8440401984480346, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.997481346130371, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8728404641151428, "num_tokens": 552874136.0, "step": 14496 }, { "epoch": 1.844167408726625, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.959654808044434, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8674660921096802, "num_tokens": 552912314.0, "step": 14497 }, { "epoch": 1.8442946190052156, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.016693115234375, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8577420115470886, "num_tokens": 552953998.0, "step": 14498 }, { "epoch": 1.8444218292838062, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.011165618896484, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.889187216758728, "num_tokens": 552987310.0, "step": 14499 }, { "epoch": 1.8445490395623967, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.01296615600586, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8683643341064453, "num_tokens": 553019830.0, "step": 14500 }, { "epoch": 1.8446762498409872, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 12.968802452087402, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8690800666809082, "num_tokens": 553066695.0, "step": 14501 }, { "epoch": 1.8448034601195777, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.044382095336914, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8623067140579224, "num_tokens": 553098943.0, "step": 14502 }, { "epoch": 1.8449306703981683, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.007148742675781, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8741239309310913, "num_tokens": 553136106.0, "step": 14503 }, { "epoch": 1.8450578806767588, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.946685791015625, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8771353363990784, "num_tokens": 553170014.0, "step": 14504 }, { "epoch": 1.8451850909553493, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.053466796875, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8644572496414185, "num_tokens": 553213297.0, "step": 14505 }, { "epoch": 1.8453123012339399, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 12.973784446716309, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8712596297264099, "num_tokens": 553248559.0, "step": 14506 }, { "epoch": 1.8454395115125302, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.079139709472656, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8720123767852783, "num_tokens": 553289140.0, "step": 14507 }, { "epoch": 1.8455667217911207, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.964241027832031, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8626843690872192, "num_tokens": 553330619.0, "step": 14508 }, { "epoch": 1.8456939320697112, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.033183097839355, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.881441593170166, "num_tokens": 553371295.0, "step": 14509 }, { "epoch": 1.8458211423483017, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.05833625793457, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8668539524078369, "num_tokens": 553410237.0, "step": 14510 }, { "epoch": 1.8459483526268923, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.03073787689209, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8776982426643372, "num_tokens": 553446727.0, "step": 14511 }, { "epoch": 1.8460755629054828, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.949787139892578, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8762649297714233, "num_tokens": 553488626.0, "step": 14512 }, { "epoch": 1.846202773184073, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.995687484741211, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8634558320045471, "num_tokens": 553528294.0, "step": 14513 }, { "epoch": 1.8463299834626636, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.004411697387695, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8745635747909546, "num_tokens": 553563738.0, "step": 14514 }, { "epoch": 1.8464571937412542, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.063810348510742, "learning_rate": 1e-06, "loss": 0.518, "mean_token_accuracy": 0.8386160135269165, "num_tokens": 553603351.0, "step": 14515 }, { "epoch": 1.8465844040198447, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.020923614501953, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8685846328735352, "num_tokens": 553643655.0, "step": 14516 }, { "epoch": 1.8467116142984352, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.997567176818848, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8820993900299072, "num_tokens": 553677944.0, "step": 14517 }, { "epoch": 1.8468388245770258, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.003838539123535, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8665986061096191, "num_tokens": 553712120.0, "step": 14518 }, { "epoch": 1.8469660348556163, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.981642723083496, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8737137317657471, "num_tokens": 553751442.0, "step": 14519 }, { "epoch": 1.8470932451342068, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.971297264099121, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8765943050384521, "num_tokens": 553782924.0, "step": 14520 }, { "epoch": 1.8472204554127973, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.029901504516602, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8613479733467102, "num_tokens": 553818861.0, "step": 14521 }, { "epoch": 1.8473476656913879, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.050185203552246, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8701902627944946, "num_tokens": 553855593.0, "step": 14522 }, { "epoch": 1.8474748759699784, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.003851890563965, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8686187267303467, "num_tokens": 553896028.0, "step": 14523 }, { "epoch": 1.847602086248569, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.019535064697266, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8742157220840454, "num_tokens": 553935212.0, "step": 14524 }, { "epoch": 1.8477292965271594, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.989828109741211, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8688820600509644, "num_tokens": 553971231.0, "step": 14525 }, { "epoch": 1.84785650680575, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.031647682189941, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8780680298805237, "num_tokens": 554006782.0, "step": 14526 }, { "epoch": 1.8479837170843405, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.087825775146484, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8625919818878174, "num_tokens": 554043427.0, "step": 14527 }, { "epoch": 1.848110927362931, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.977112770080566, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8600871562957764, "num_tokens": 554084187.0, "step": 14528 }, { "epoch": 1.8482381376415216, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.062594413757324, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8704344630241394, "num_tokens": 554122186.0, "step": 14529 }, { "epoch": 1.848365347920112, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.988920211791992, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8700205683708191, "num_tokens": 554161388.0, "step": 14530 }, { "epoch": 1.8484925581987026, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.083269119262695, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8620554804801941, "num_tokens": 554199837.0, "step": 14531 }, { "epoch": 1.848619768477293, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.088767051696777, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8605391979217529, "num_tokens": 554239567.0, "step": 14532 }, { "epoch": 1.8487469787558835, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.034998893737793, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.873801589012146, "num_tokens": 554277040.0, "step": 14533 }, { "epoch": 1.848874189034474, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.990469932556152, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8743590712547302, "num_tokens": 554318994.0, "step": 14534 }, { "epoch": 1.8490013993130645, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.075312614440918, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8636517524719238, "num_tokens": 554355334.0, "step": 14535 }, { "epoch": 1.849128609591655, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.968489646911621, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8592645525932312, "num_tokens": 554396504.0, "step": 14536 }, { "epoch": 1.8492558198702456, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.034517288208008, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8767470121383667, "num_tokens": 554429772.0, "step": 14537 }, { "epoch": 1.8493830301488359, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.033257484436035, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8722305297851562, "num_tokens": 554470102.0, "step": 14538 }, { "epoch": 1.8495102404274264, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.996295928955078, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8613306283950806, "num_tokens": 554511632.0, "step": 14539 }, { "epoch": 1.849637450706017, "ewc_loss": 0.031982421875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 13.007661819458008, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8661884665489197, "num_tokens": 554553028.0, "step": 14540 }, { "epoch": 1.8497646609846075, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.032902717590332, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8644150495529175, "num_tokens": 554585109.0, "step": 14541 }, { "epoch": 1.849891871263198, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.97507095336914, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.870855450630188, "num_tokens": 554621141.0, "step": 14542 }, { "epoch": 1.8500190815417885, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.976856231689453, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.869587779045105, "num_tokens": 554660838.0, "step": 14543 }, { "epoch": 1.850146291820379, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 12.996232032775879, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8717544674873352, "num_tokens": 554696338.0, "step": 14544 }, { "epoch": 1.8502735020989696, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.00252628326416, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8824474811553955, "num_tokens": 554729782.0, "step": 14545 }, { "epoch": 1.85040071237756, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.070844650268555, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8857261538505554, "num_tokens": 554767566.0, "step": 14546 }, { "epoch": 1.8505279226561506, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.98755931854248, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8750287294387817, "num_tokens": 554806063.0, "step": 14547 }, { "epoch": 1.8506551329347412, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.051865577697754, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8663095235824585, "num_tokens": 554839418.0, "step": 14548 }, { "epoch": 1.8507823432133317, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.018697738647461, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8657456040382385, "num_tokens": 554881888.0, "step": 14549 }, { "epoch": 1.8509095534919222, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.995241165161133, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8887594938278198, "num_tokens": 554921561.0, "step": 14550 }, { "epoch": 1.8510367637705127, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.006248474121094, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8468878269195557, "num_tokens": 554954380.0, "step": 14551 }, { "epoch": 1.8511639740491033, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.087969779968262, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8709014058113098, "num_tokens": 554994781.0, "step": 14552 }, { "epoch": 1.8512911843276938, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.063680648803711, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8784821629524231, "num_tokens": 555024938.0, "step": 14553 }, { "epoch": 1.8514183946062843, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.061644554138184, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8590734004974365, "num_tokens": 555059785.0, "step": 14554 }, { "epoch": 1.8515456048848749, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.983155250549316, "learning_rate": 1e-06, "loss": 0.4821, "mean_token_accuracy": 0.844470739364624, "num_tokens": 555097502.0, "step": 14555 }, { "epoch": 1.8516728151634652, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.026345252990723, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8691499829292297, "num_tokens": 555139976.0, "step": 14556 }, { "epoch": 1.8518000254420557, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.970571517944336, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8726603388786316, "num_tokens": 555181250.0, "step": 14557 }, { "epoch": 1.8519272357206462, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.02811050415039, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.8512065410614014, "num_tokens": 555214927.0, "step": 14558 }, { "epoch": 1.8520544459992367, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.968977928161621, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8667605519294739, "num_tokens": 555255795.0, "step": 14559 }, { "epoch": 1.8521816562778273, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.034177780151367, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8728752136230469, "num_tokens": 555294177.0, "step": 14560 }, { "epoch": 1.8523088665564178, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.998346328735352, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8822221755981445, "num_tokens": 555331548.0, "step": 14561 }, { "epoch": 1.852436076835008, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 12.991059303283691, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8824528455734253, "num_tokens": 555368426.0, "step": 14562 }, { "epoch": 1.8525632871135986, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.004341125488281, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.886948823928833, "num_tokens": 555411594.0, "step": 14563 }, { "epoch": 1.8526904973921892, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.022268295288086, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8617424964904785, "num_tokens": 555450412.0, "step": 14564 }, { "epoch": 1.8528177076707797, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.024211883544922, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8785760402679443, "num_tokens": 555489067.0, "step": 14565 }, { "epoch": 1.8529449179493702, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.98134994506836, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8768855929374695, "num_tokens": 555524552.0, "step": 14566 }, { "epoch": 1.8530721282279607, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.975282669067383, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8440225124359131, "num_tokens": 555563766.0, "step": 14567 }, { "epoch": 1.8531993385065513, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.036426544189453, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8768570423126221, "num_tokens": 555601410.0, "step": 14568 }, { "epoch": 1.8533265487851418, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.039456367492676, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8700820207595825, "num_tokens": 555641346.0, "step": 14569 }, { "epoch": 1.8534537590637323, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.042746543884277, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8797128796577454, "num_tokens": 555676537.0, "step": 14570 }, { "epoch": 1.8535809693423229, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.019781112670898, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.855647087097168, "num_tokens": 555721634.0, "step": 14571 }, { "epoch": 1.8537081796209134, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.985428810119629, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8696409463882446, "num_tokens": 555765447.0, "step": 14572 }, { "epoch": 1.853835389899504, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.102395057678223, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8695781826972961, "num_tokens": 555798598.0, "step": 14573 }, { "epoch": 1.8539626001780944, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.020163536071777, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8798853158950806, "num_tokens": 555829465.0, "step": 14574 }, { "epoch": 1.854089810456685, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.043294906616211, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8712658882141113, "num_tokens": 555866577.0, "step": 14575 }, { "epoch": 1.8542170207352755, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.042959213256836, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.873294472694397, "num_tokens": 555900472.0, "step": 14576 }, { "epoch": 1.854344231013866, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.041476249694824, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8653327226638794, "num_tokens": 555936356.0, "step": 14577 }, { "epoch": 1.8544714412924566, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.072978973388672, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.863643229007721, "num_tokens": 555975338.0, "step": 14578 }, { "epoch": 1.854598651571047, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.997492790222168, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8610843420028687, "num_tokens": 556019651.0, "step": 14579 }, { "epoch": 1.8547258618496376, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.073744773864746, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8655940294265747, "num_tokens": 556054530.0, "step": 14580 }, { "epoch": 1.854853072128228, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.023143768310547, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8658512234687805, "num_tokens": 556091985.0, "step": 14581 }, { "epoch": 1.8549802824068184, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.998698234558105, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8722876906394958, "num_tokens": 556128777.0, "step": 14582 }, { "epoch": 1.855107492685409, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.995197296142578, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.8890464305877686, "num_tokens": 556164286.0, "step": 14583 }, { "epoch": 1.8552347029639995, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.982873916625977, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8778321146965027, "num_tokens": 556199141.0, "step": 14584 }, { "epoch": 1.85536191324259, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.069283485412598, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8748763799667358, "num_tokens": 556237104.0, "step": 14585 }, { "epoch": 1.8554891235211806, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 12.982524871826172, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8592902421951294, "num_tokens": 556275343.0, "step": 14586 }, { "epoch": 1.8556163337997709, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.00121021270752, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8854501843452454, "num_tokens": 556316539.0, "step": 14587 }, { "epoch": 1.8557435440783614, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.048721313476562, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8827086687088013, "num_tokens": 556351101.0, "step": 14588 }, { "epoch": 1.855870754356952, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.980120658874512, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.883311927318573, "num_tokens": 556382458.0, "step": 14589 }, { "epoch": 1.8559979646355425, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.053080558776855, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8725845813751221, "num_tokens": 556416980.0, "step": 14590 }, { "epoch": 1.856125174914133, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.031659126281738, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8766376972198486, "num_tokens": 556451868.0, "step": 14591 }, { "epoch": 1.8562523851927235, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.991244316101074, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8586316108703613, "num_tokens": 556485682.0, "step": 14592 }, { "epoch": 1.856379595471314, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.035655975341797, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8816393613815308, "num_tokens": 556524857.0, "step": 14593 }, { "epoch": 1.8565068057499046, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.010995864868164, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8748574256896973, "num_tokens": 556565763.0, "step": 14594 }, { "epoch": 1.856634016028495, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.996016502380371, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8589955568313599, "num_tokens": 556606202.0, "step": 14595 }, { "epoch": 1.8567612263070856, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.05525016784668, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8798976540565491, "num_tokens": 556646906.0, "step": 14596 }, { "epoch": 1.8568884365856761, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 12.996718406677246, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8729601502418518, "num_tokens": 556684249.0, "step": 14597 }, { "epoch": 1.8570156468642667, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.034729957580566, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8602105379104614, "num_tokens": 556726840.0, "step": 14598 }, { "epoch": 1.8571428571428572, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.027412414550781, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8742735385894775, "num_tokens": 556764822.0, "step": 14599 }, { "epoch": 1.8572700674214477, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.055170059204102, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8687268495559692, "num_tokens": 556800884.0, "step": 14600 }, { "epoch": 1.8573972777000383, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.021661758422852, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8765703439712524, "num_tokens": 556841233.0, "step": 14601 }, { "epoch": 1.8575244879786288, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.049840927124023, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8648988008499146, "num_tokens": 556880995.0, "step": 14602 }, { "epoch": 1.8576516982572193, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.05195426940918, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8758698105812073, "num_tokens": 556914367.0, "step": 14603 }, { "epoch": 1.8577789085358098, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.067242622375488, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8515132069587708, "num_tokens": 556955944.0, "step": 14604 }, { "epoch": 1.8579061188144002, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.060574531555176, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8541298508644104, "num_tokens": 556989851.0, "step": 14605 }, { "epoch": 1.8580333290929907, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.071396827697754, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8704447746276855, "num_tokens": 557033662.0, "step": 14606 }, { "epoch": 1.8581605393715812, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 12.951554298400879, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8637263178825378, "num_tokens": 557074869.0, "step": 14607 }, { "epoch": 1.8582877496501717, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.084035873413086, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8606167435646057, "num_tokens": 557110164.0, "step": 14608 }, { "epoch": 1.8584149599287623, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.007038116455078, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8708682060241699, "num_tokens": 557149309.0, "step": 14609 }, { "epoch": 1.8585421702073528, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.083717346191406, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8659995198249817, "num_tokens": 557180510.0, "step": 14610 }, { "epoch": 1.858669380485943, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.025487899780273, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8761132955551147, "num_tokens": 557218062.0, "step": 14611 }, { "epoch": 1.8587965907645336, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.013577461242676, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8626458048820496, "num_tokens": 557258963.0, "step": 14612 }, { "epoch": 1.8589238010431242, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.05565071105957, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8713394403457642, "num_tokens": 557296046.0, "step": 14613 }, { "epoch": 1.8590510113217147, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.051825523376465, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8679111003875732, "num_tokens": 557328395.0, "step": 14614 }, { "epoch": 1.8591782216003052, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.031660079956055, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8729698061943054, "num_tokens": 557363888.0, "step": 14615 }, { "epoch": 1.8593054318788957, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.992162704467773, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8634519577026367, "num_tokens": 557402944.0, "step": 14616 }, { "epoch": 1.8594326421574863, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.04001235961914, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8765722513198853, "num_tokens": 557442247.0, "step": 14617 }, { "epoch": 1.8595598524360768, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.063311576843262, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8649590611457825, "num_tokens": 557483580.0, "step": 14618 }, { "epoch": 1.8596870627146673, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.070301055908203, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8743904829025269, "num_tokens": 557520134.0, "step": 14619 }, { "epoch": 1.8598142729932579, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.022595405578613, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8651356101036072, "num_tokens": 557565726.0, "step": 14620 }, { "epoch": 1.8599414832718484, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.056012153625488, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.869542121887207, "num_tokens": 557602871.0, "step": 14621 }, { "epoch": 1.860068693550439, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.05742073059082, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8614209890365601, "num_tokens": 557640873.0, "step": 14622 }, { "epoch": 1.8601959038290294, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.053276062011719, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.86346036195755, "num_tokens": 557684233.0, "step": 14623 }, { "epoch": 1.86032311410762, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.055745124816895, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8744533061981201, "num_tokens": 557716895.0, "step": 14624 }, { "epoch": 1.8604503243862105, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.039517402648926, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8671157360076904, "num_tokens": 557758197.0, "step": 14625 }, { "epoch": 1.860577534664801, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.064505577087402, "learning_rate": 1e-06, "loss": 0.5076, "mean_token_accuracy": 0.8428956866264343, "num_tokens": 557794196.0, "step": 14626 }, { "epoch": 1.8607047449433916, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.079333305358887, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8540766835212708, "num_tokens": 557833856.0, "step": 14627 }, { "epoch": 1.860831955221982, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.013888359069824, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8586689233779907, "num_tokens": 557866304.0, "step": 14628 }, { "epoch": 1.8609591655005726, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.05301570892334, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8825616240501404, "num_tokens": 557907065.0, "step": 14629 }, { "epoch": 1.861086375779163, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.037394523620605, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8654302358627319, "num_tokens": 557947326.0, "step": 14630 }, { "epoch": 1.8612135860577534, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.044178009033203, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.868444561958313, "num_tokens": 557990059.0, "step": 14631 }, { "epoch": 1.861340796336344, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.022919654846191, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8720029592514038, "num_tokens": 558028406.0, "step": 14632 }, { "epoch": 1.8614680066149345, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.076852798461914, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8651931285858154, "num_tokens": 558070239.0, "step": 14633 }, { "epoch": 1.861595216893525, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.08342456817627, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8739480376243591, "num_tokens": 558107685.0, "step": 14634 }, { "epoch": 1.8617224271721156, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.056770324707031, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8631543517112732, "num_tokens": 558143185.0, "step": 14635 }, { "epoch": 1.8618496374507059, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.077056884765625, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.878699004650116, "num_tokens": 558176207.0, "step": 14636 }, { "epoch": 1.8619768477292964, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.08146858215332, "learning_rate": 1e-06, "loss": 0.3545, "mean_token_accuracy": 0.8859589099884033, "num_tokens": 558212215.0, "step": 14637 }, { "epoch": 1.862104058007887, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.06689167022705, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8793982863426208, "num_tokens": 558248471.0, "step": 14638 }, { "epoch": 1.8622312682864774, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.132009506225586, "learning_rate": 1e-06, "loss": 0.4523, "mean_token_accuracy": 0.8576084971427917, "num_tokens": 558285211.0, "step": 14639 }, { "epoch": 1.862358478565068, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.109853744506836, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.863127589225769, "num_tokens": 558314434.0, "step": 14640 }, { "epoch": 1.8624856888436585, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.090838432312012, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8778194189071655, "num_tokens": 558353710.0, "step": 14641 }, { "epoch": 1.862612899122249, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.124651908874512, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.85726398229599, "num_tokens": 558390476.0, "step": 14642 }, { "epoch": 1.8627401094008396, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.00565242767334, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8788073062896729, "num_tokens": 558424917.0, "step": 14643 }, { "epoch": 1.86286731967943, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.090553283691406, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8642975091934204, "num_tokens": 558459831.0, "step": 14644 }, { "epoch": 1.8629945299580206, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.068184852600098, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8844580054283142, "num_tokens": 558498600.0, "step": 14645 }, { "epoch": 1.8631217402366111, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.030482292175293, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8608186841011047, "num_tokens": 558537465.0, "step": 14646 }, { "epoch": 1.8632489505152017, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.107935905456543, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8897570371627808, "num_tokens": 558575357.0, "step": 14647 }, { "epoch": 1.8633761607937922, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.098616600036621, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8671277761459351, "num_tokens": 558619422.0, "step": 14648 }, { "epoch": 1.8635033710723827, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.032482147216797, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.861860454082489, "num_tokens": 558657677.0, "step": 14649 }, { "epoch": 1.8636305813509733, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.081497192382812, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8779710531234741, "num_tokens": 558692912.0, "step": 14650 }, { "epoch": 1.8637577916295638, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.094103813171387, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8818562030792236, "num_tokens": 558733470.0, "step": 14651 }, { "epoch": 1.8638850019081543, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.025321960449219, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8662235736846924, "num_tokens": 558771899.0, "step": 14652 }, { "epoch": 1.8640122121867448, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.08718204498291, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8759996294975281, "num_tokens": 558813002.0, "step": 14653 }, { "epoch": 1.8641394224653351, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.014678001403809, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8549845218658447, "num_tokens": 558850827.0, "step": 14654 }, { "epoch": 1.8642666327439257, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.020773887634277, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8669925928115845, "num_tokens": 558889003.0, "step": 14655 }, { "epoch": 1.8643938430225162, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.050420761108398, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8766822218894958, "num_tokens": 558925849.0, "step": 14656 }, { "epoch": 1.8645210533011067, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 12.995016098022461, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8693736791610718, "num_tokens": 558967471.0, "step": 14657 }, { "epoch": 1.8646482635796973, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.04932689666748, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8736875653266907, "num_tokens": 559001938.0, "step": 14658 }, { "epoch": 1.8647754738582878, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.016813278198242, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8756895661354065, "num_tokens": 559040515.0, "step": 14659 }, { "epoch": 1.864902684136878, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.0439453125, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.88205885887146, "num_tokens": 559077702.0, "step": 14660 }, { "epoch": 1.8650298944154686, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.0143461227417, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.863628625869751, "num_tokens": 559114741.0, "step": 14661 }, { "epoch": 1.8651571046940592, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.082464218139648, "learning_rate": 1e-06, "loss": 0.4836, "mean_token_accuracy": 0.8486768007278442, "num_tokens": 559155844.0, "step": 14662 }, { "epoch": 1.8652843149726497, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.039244651794434, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8721281886100769, "num_tokens": 559194964.0, "step": 14663 }, { "epoch": 1.8654115252512402, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.047464370727539, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8759549856185913, "num_tokens": 559230083.0, "step": 14664 }, { "epoch": 1.8655387355298307, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.023439407348633, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8809629678726196, "num_tokens": 559268397.0, "step": 14665 }, { "epoch": 1.8656659458084213, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.083024978637695, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8618025779724121, "num_tokens": 559302024.0, "step": 14666 }, { "epoch": 1.8657931560870118, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.08828067779541, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8674464821815491, "num_tokens": 559343192.0, "step": 14667 }, { "epoch": 1.8659203663656023, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.04920768737793, "learning_rate": 1e-06, "loss": 0.3401, "mean_token_accuracy": 0.8918083906173706, "num_tokens": 559380616.0, "step": 14668 }, { "epoch": 1.8660475766441929, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.113462448120117, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.880500316619873, "num_tokens": 559413621.0, "step": 14669 }, { "epoch": 1.8661747869227834, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.056527137756348, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8792912364006042, "num_tokens": 559447274.0, "step": 14670 }, { "epoch": 1.866301997201374, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.082555770874023, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8793838024139404, "num_tokens": 559484306.0, "step": 14671 }, { "epoch": 1.8664292074799644, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 12.992302894592285, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8763115406036377, "num_tokens": 559527126.0, "step": 14672 }, { "epoch": 1.866556417758555, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.020160675048828, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8743279576301575, "num_tokens": 559568342.0, "step": 14673 }, { "epoch": 1.8666836280371455, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.093438148498535, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8681997060775757, "num_tokens": 559609597.0, "step": 14674 }, { "epoch": 1.866810838315736, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.081860542297363, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8685241937637329, "num_tokens": 559646392.0, "step": 14675 }, { "epoch": 1.8669380485943265, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.108972549438477, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8739787340164185, "num_tokens": 559684836.0, "step": 14676 }, { "epoch": 1.867065258872917, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.047900199890137, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8749086856842041, "num_tokens": 559727793.0, "step": 14677 }, { "epoch": 1.8671924691515076, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.096715927124023, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8596925735473633, "num_tokens": 559768154.0, "step": 14678 }, { "epoch": 1.867319679430098, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.029706954956055, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8719229698181152, "num_tokens": 559798732.0, "step": 14679 }, { "epoch": 1.8674468897086884, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.045578002929688, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8696287274360657, "num_tokens": 559842796.0, "step": 14680 }, { "epoch": 1.867574099987279, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.049341201782227, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.867555558681488, "num_tokens": 559884198.0, "step": 14681 }, { "epoch": 1.8677013102658695, "ewc_loss": 0.0322265625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.111093521118164, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8567153215408325, "num_tokens": 559916749.0, "step": 14682 }, { "epoch": 1.86782852054446, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.030282020568848, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8659061193466187, "num_tokens": 559958430.0, "step": 14683 }, { "epoch": 1.8679557308230506, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.050247192382812, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8871548771858215, "num_tokens": 559995720.0, "step": 14684 }, { "epoch": 1.8680829411016409, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.039682388305664, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.873967707157135, "num_tokens": 560033304.0, "step": 14685 }, { "epoch": 1.8682101513802314, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.085310935974121, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8692601919174194, "num_tokens": 560069879.0, "step": 14686 }, { "epoch": 1.868337361658822, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.083549499511719, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8686789870262146, "num_tokens": 560106853.0, "step": 14687 }, { "epoch": 1.8684645719374124, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.056310653686523, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.87246173620224, "num_tokens": 560144225.0, "step": 14688 }, { "epoch": 1.868591782216003, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.028681755065918, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8644853830337524, "num_tokens": 560186520.0, "step": 14689 }, { "epoch": 1.8687189924945935, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.123220443725586, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8737922310829163, "num_tokens": 560225458.0, "step": 14690 }, { "epoch": 1.868846202773184, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.029239654541016, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.874985933303833, "num_tokens": 560263401.0, "step": 14691 }, { "epoch": 1.8689734130517746, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.063974380493164, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8678948879241943, "num_tokens": 560305026.0, "step": 14692 }, { "epoch": 1.869100623330365, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.072157859802246, "learning_rate": 1e-06, "loss": 0.5075, "mean_token_accuracy": 0.8386526107788086, "num_tokens": 560354823.0, "step": 14693 }, { "epoch": 1.8692278336089556, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.03806209564209, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8849357962608337, "num_tokens": 560391888.0, "step": 14694 }, { "epoch": 1.8693550438875461, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.088574409484863, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8731093406677246, "num_tokens": 560425350.0, "step": 14695 }, { "epoch": 1.8694822541661367, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.029284477233887, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8809067010879517, "num_tokens": 560458521.0, "step": 14696 }, { "epoch": 1.8696094644447272, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.077960014343262, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8651774525642395, "num_tokens": 560506378.0, "step": 14697 }, { "epoch": 1.8697366747233177, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.112064361572266, "learning_rate": 1e-06, "loss": 0.3357, "mean_token_accuracy": 0.8927297592163086, "num_tokens": 560544055.0, "step": 14698 }, { "epoch": 1.8698638850019083, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.039566993713379, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8800510764122009, "num_tokens": 560577240.0, "step": 14699 }, { "epoch": 1.8699910952804988, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.11075210571289, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8660504817962646, "num_tokens": 560615662.0, "step": 14700 }, { "epoch": 1.8701183055590893, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.03465461730957, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8662168979644775, "num_tokens": 560658386.0, "step": 14701 }, { "epoch": 1.8702455158376798, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.078901290893555, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8687071204185486, "num_tokens": 560696366.0, "step": 14702 }, { "epoch": 1.8703727261162701, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.074113845825195, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8600080013275146, "num_tokens": 560736419.0, "step": 14703 }, { "epoch": 1.8704999363948607, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.089247703552246, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8695820569992065, "num_tokens": 560777839.0, "step": 14704 }, { "epoch": 1.8706271466734512, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.07016372680664, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8810378313064575, "num_tokens": 560816167.0, "step": 14705 }, { "epoch": 1.8707543569520417, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.0950345993042, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8569937944412231, "num_tokens": 560854524.0, "step": 14706 }, { "epoch": 1.8708815672306323, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.088814735412598, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8675584197044373, "num_tokens": 560894393.0, "step": 14707 }, { "epoch": 1.8710087775092228, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.053895950317383, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.880874752998352, "num_tokens": 560938640.0, "step": 14708 }, { "epoch": 1.871135987787813, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.092667579650879, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8695300817489624, "num_tokens": 560976805.0, "step": 14709 }, { "epoch": 1.8712631980664036, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.058467864990234, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8682329654693604, "num_tokens": 561018940.0, "step": 14710 }, { "epoch": 1.8713904083449941, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.028691291809082, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8756291270256042, "num_tokens": 561065228.0, "step": 14711 }, { "epoch": 1.8715176186235847, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.079378128051758, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8625079989433289, "num_tokens": 561107540.0, "step": 14712 }, { "epoch": 1.8716448289021752, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.121121406555176, "learning_rate": 1e-06, "loss": 0.4959, "mean_token_accuracy": 0.8512948751449585, "num_tokens": 561144774.0, "step": 14713 }, { "epoch": 1.8717720391807657, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.0700044631958, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8513701558113098, "num_tokens": 561186774.0, "step": 14714 }, { "epoch": 1.8718992494593563, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 13.120616912841797, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8831652998924255, "num_tokens": 561223997.0, "step": 14715 }, { "epoch": 1.8720264597379468, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.081945419311523, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8780395984649658, "num_tokens": 561261583.0, "step": 14716 }, { "epoch": 1.8721536700165373, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.077320098876953, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8752436637878418, "num_tokens": 561302681.0, "step": 14717 }, { "epoch": 1.8722808802951278, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.086912155151367, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8692404627799988, "num_tokens": 561338402.0, "step": 14718 }, { "epoch": 1.8724080905737184, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.127114295959473, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8741170167922974, "num_tokens": 561372892.0, "step": 14719 }, { "epoch": 1.872535300852309, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.073716163635254, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8676366806030273, "num_tokens": 561409474.0, "step": 14720 }, { "epoch": 1.8726625111308994, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.178024291992188, "learning_rate": 1e-06, "loss": 0.4866, "mean_token_accuracy": 0.8445605039596558, "num_tokens": 561450152.0, "step": 14721 }, { "epoch": 1.87278972140949, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.065290451049805, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8639264702796936, "num_tokens": 561489448.0, "step": 14722 }, { "epoch": 1.8729169316880805, "ewc_loss": 0.032470703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 13.084778785705566, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.8477885723114014, "num_tokens": 561529684.0, "step": 14723 }, { "epoch": 1.873044141966671, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.121853828430176, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8715391159057617, "num_tokens": 561564259.0, "step": 14724 }, { "epoch": 1.8731713522452615, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.017243385314941, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8783596754074097, "num_tokens": 561609650.0, "step": 14725 }, { "epoch": 1.873298562523852, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.096220016479492, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.876329243183136, "num_tokens": 561645520.0, "step": 14726 }, { "epoch": 1.8734257728024426, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.121700286865234, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8754281997680664, "num_tokens": 561682771.0, "step": 14727 }, { "epoch": 1.873552983081033, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.105311393737793, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8603416681289673, "num_tokens": 561721317.0, "step": 14728 }, { "epoch": 1.8736801933596234, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.100610733032227, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8716394901275635, "num_tokens": 561763178.0, "step": 14729 }, { "epoch": 1.873807403638214, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.125499725341797, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8711228370666504, "num_tokens": 561802292.0, "step": 14730 }, { "epoch": 1.8739346139168045, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.131368637084961, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8784757852554321, "num_tokens": 561838610.0, "step": 14731 }, { "epoch": 1.874061824195395, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.109831809997559, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8747602701187134, "num_tokens": 561881992.0, "step": 14732 }, { "epoch": 1.8741890344739855, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.156506538391113, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8614832162857056, "num_tokens": 561917661.0, "step": 14733 }, { "epoch": 1.8743162447525759, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.103960037231445, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8721504211425781, "num_tokens": 561958218.0, "step": 14734 }, { "epoch": 1.8744434550311664, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.126313209533691, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8819136023521423, "num_tokens": 561993882.0, "step": 14735 }, { "epoch": 1.874570665309757, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.107959747314453, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8715394735336304, "num_tokens": 562035443.0, "step": 14736 }, { "epoch": 1.8746978755883474, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.106913566589355, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8681197166442871, "num_tokens": 562072806.0, "step": 14737 }, { "epoch": 1.874825085866938, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.107653617858887, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8538649082183838, "num_tokens": 562113652.0, "step": 14738 }, { "epoch": 1.8749522961455285, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.112204551696777, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8628053069114685, "num_tokens": 562150600.0, "step": 14739 }, { "epoch": 1.875079506424119, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.110114097595215, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.872331976890564, "num_tokens": 562187730.0, "step": 14740 }, { "epoch": 1.8752067167027096, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.059383392333984, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8701969385147095, "num_tokens": 562228229.0, "step": 14741 }, { "epoch": 1.8753339269813, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.19052505493164, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8765453100204468, "num_tokens": 562262293.0, "step": 14742 }, { "epoch": 1.8754611372598906, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.082494735717773, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8677496910095215, "num_tokens": 562303389.0, "step": 14743 }, { "epoch": 1.8755883475384811, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.095423698425293, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8643503785133362, "num_tokens": 562344461.0, "step": 14744 }, { "epoch": 1.8757155578170717, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.109843254089355, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.853318452835083, "num_tokens": 562387068.0, "step": 14745 }, { "epoch": 1.8758427680956622, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.113317489624023, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8683536052703857, "num_tokens": 562428225.0, "step": 14746 }, { "epoch": 1.8759699783742527, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.133623123168945, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8704536557197571, "num_tokens": 562463879.0, "step": 14747 }, { "epoch": 1.8760971886528433, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.09190559387207, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8635175228118896, "num_tokens": 562502327.0, "step": 14748 }, { "epoch": 1.8762243989314338, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.13770580291748, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8757650852203369, "num_tokens": 562542162.0, "step": 14749 }, { "epoch": 1.8763516092100243, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.105100631713867, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8652679324150085, "num_tokens": 562583364.0, "step": 14750 }, { "epoch": 1.8764788194886148, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.1216402053833, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8727632761001587, "num_tokens": 562615870.0, "step": 14751 }, { "epoch": 1.8766060297672051, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.125432014465332, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8555078506469727, "num_tokens": 562655086.0, "step": 14752 }, { "epoch": 1.8767332400457957, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.107573509216309, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8630267977714539, "num_tokens": 562697532.0, "step": 14753 }, { "epoch": 1.8768604503243862, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.11349868774414, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8504536747932434, "num_tokens": 562731057.0, "step": 14754 }, { "epoch": 1.8769876606029767, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.136310577392578, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8591041564941406, "num_tokens": 562770402.0, "step": 14755 }, { "epoch": 1.8771148708815673, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.052681922912598, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.867215633392334, "num_tokens": 562806191.0, "step": 14756 }, { "epoch": 1.8772420811601578, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.154169082641602, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8690502643585205, "num_tokens": 562846864.0, "step": 14757 }, { "epoch": 1.877369291438748, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.068626403808594, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8647763729095459, "num_tokens": 562882887.0, "step": 14758 }, { "epoch": 1.8774965017173386, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.085542678833008, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8689727783203125, "num_tokens": 562919096.0, "step": 14759 }, { "epoch": 1.8776237119959291, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.109521865844727, "learning_rate": 1e-06, "loss": 0.4969, "mean_token_accuracy": 0.8433266282081604, "num_tokens": 562959336.0, "step": 14760 }, { "epoch": 1.8777509222745197, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.123818397521973, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8687273859977722, "num_tokens": 562997883.0, "step": 14761 }, { "epoch": 1.8778781325531102, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.089675903320312, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8787796497344971, "num_tokens": 563036763.0, "step": 14762 }, { "epoch": 1.8780053428317007, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.096572875976562, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.886091947555542, "num_tokens": 563076815.0, "step": 14763 }, { "epoch": 1.8781325531102913, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.079197883605957, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.880122184753418, "num_tokens": 563110511.0, "step": 14764 }, { "epoch": 1.8782597633888818, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.140949249267578, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8640148639678955, "num_tokens": 563140757.0, "step": 14765 }, { "epoch": 1.8783869736674723, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.043120384216309, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8693773746490479, "num_tokens": 563179961.0, "step": 14766 }, { "epoch": 1.8785141839460628, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.110434532165527, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8768296241760254, "num_tokens": 563221355.0, "step": 14767 }, { "epoch": 1.8786413942246534, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.080351829528809, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.890383243560791, "num_tokens": 563257427.0, "step": 14768 }, { "epoch": 1.878768604503244, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.055017471313477, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8667286038398743, "num_tokens": 563294371.0, "step": 14769 }, { "epoch": 1.8788958147818344, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.137578964233398, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8798428773880005, "num_tokens": 563338592.0, "step": 14770 }, { "epoch": 1.879023025060425, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.091679573059082, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8736008405685425, "num_tokens": 563373853.0, "step": 14771 }, { "epoch": 1.8791502353390155, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.088723182678223, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8696954846382141, "num_tokens": 563411137.0, "step": 14772 }, { "epoch": 1.879277445617606, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.090636253356934, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8735699653625488, "num_tokens": 563450294.0, "step": 14773 }, { "epoch": 1.8794046558961965, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.054426193237305, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8712234497070312, "num_tokens": 563496858.0, "step": 14774 }, { "epoch": 1.879531866174787, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.12778377532959, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.865790843963623, "num_tokens": 563539566.0, "step": 14775 }, { "epoch": 1.8796590764533776, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.133708953857422, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8783814907073975, "num_tokens": 563574363.0, "step": 14776 }, { "epoch": 1.879786286731968, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.143335342407227, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8720651865005493, "num_tokens": 563605540.0, "step": 14777 }, { "epoch": 1.8799134970105584, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.129511833190918, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8562310934066772, "num_tokens": 563646640.0, "step": 14778 }, { "epoch": 1.880040707289149, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.056262969970703, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8807918429374695, "num_tokens": 563683576.0, "step": 14779 }, { "epoch": 1.8801679175677395, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.159244537353516, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8622547388076782, "num_tokens": 563724088.0, "step": 14780 }, { "epoch": 1.88029512784633, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.092338562011719, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8689538836479187, "num_tokens": 563760897.0, "step": 14781 }, { "epoch": 1.8804223381249205, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.107658386230469, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.875139594078064, "num_tokens": 563800158.0, "step": 14782 }, { "epoch": 1.8805495484035109, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.043173789978027, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8754150867462158, "num_tokens": 563833288.0, "step": 14783 }, { "epoch": 1.8806767586821014, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.09574031829834, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8818050622940063, "num_tokens": 563874507.0, "step": 14784 }, { "epoch": 1.880803968960692, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.104440689086914, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.869383692741394, "num_tokens": 563912452.0, "step": 14785 }, { "epoch": 1.8809311792392824, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.060924530029297, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8531341552734375, "num_tokens": 563950565.0, "step": 14786 }, { "epoch": 1.881058389517873, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.130270957946777, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8610712885856628, "num_tokens": 563984924.0, "step": 14787 }, { "epoch": 1.8811855997964635, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.087932586669922, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8762953281402588, "num_tokens": 564022887.0, "step": 14788 }, { "epoch": 1.881312810075054, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.084924697875977, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8689442276954651, "num_tokens": 564064829.0, "step": 14789 }, { "epoch": 1.8814400203536445, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.103839874267578, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8490591049194336, "num_tokens": 564100858.0, "step": 14790 }, { "epoch": 1.881567230632235, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.101844787597656, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8639478087425232, "num_tokens": 564141477.0, "step": 14791 }, { "epoch": 1.8816944409108256, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.088197708129883, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8627395033836365, "num_tokens": 564179271.0, "step": 14792 }, { "epoch": 1.8818216511894161, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.06651496887207, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8762446045875549, "num_tokens": 564224336.0, "step": 14793 }, { "epoch": 1.8819488614680067, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.102255821228027, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8645801544189453, "num_tokens": 564265157.0, "step": 14794 }, { "epoch": 1.8820760717465972, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.146778106689453, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8763519525527954, "num_tokens": 564295023.0, "step": 14795 }, { "epoch": 1.8822032820251877, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.157327651977539, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8725051879882812, "num_tokens": 564333700.0, "step": 14796 }, { "epoch": 1.8823304923037782, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.072150230407715, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8567706942558289, "num_tokens": 564374342.0, "step": 14797 }, { "epoch": 1.8824577025823688, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.130184173583984, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8716667294502258, "num_tokens": 564422877.0, "step": 14798 }, { "epoch": 1.8825849128609593, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.114047050476074, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8698046803474426, "num_tokens": 564464713.0, "step": 14799 }, { "epoch": 1.8827121231395498, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.115381240844727, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8588852882385254, "num_tokens": 564500667.0, "step": 14800 }, { "epoch": 1.8828393334181401, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.108619689941406, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8788304328918457, "num_tokens": 564541460.0, "step": 14801 }, { "epoch": 1.8829665436967307, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.095379829406738, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8626947999000549, "num_tokens": 564580983.0, "step": 14802 }, { "epoch": 1.8830937539753212, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.099658966064453, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8785630464553833, "num_tokens": 564615967.0, "step": 14803 }, { "epoch": 1.8832209642539117, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.126399993896484, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.861572265625, "num_tokens": 564653255.0, "step": 14804 }, { "epoch": 1.8833481745325023, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.110440254211426, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8679313659667969, "num_tokens": 564688420.0, "step": 14805 }, { "epoch": 1.8834753848110928, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.098773956298828, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8634506464004517, "num_tokens": 564733123.0, "step": 14806 }, { "epoch": 1.883602595089683, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.138253211975098, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8768453001976013, "num_tokens": 564767712.0, "step": 14807 }, { "epoch": 1.8837298053682736, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.086236953735352, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8670713305473328, "num_tokens": 564810663.0, "step": 14808 }, { "epoch": 1.8838570156468641, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.199383735656738, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8686065077781677, "num_tokens": 564848689.0, "step": 14809 }, { "epoch": 1.8839842259254547, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.098150253295898, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8581901788711548, "num_tokens": 564887426.0, "step": 14810 }, { "epoch": 1.8841114362040452, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.082950592041016, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8739885091781616, "num_tokens": 564924852.0, "step": 14811 }, { "epoch": 1.8842386464826357, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.141579627990723, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8659108281135559, "num_tokens": 564964201.0, "step": 14812 }, { "epoch": 1.8843658567612263, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.084409713745117, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8809281587600708, "num_tokens": 565004939.0, "step": 14813 }, { "epoch": 1.8844930670398168, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.139893531799316, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.874505877494812, "num_tokens": 565044067.0, "step": 14814 }, { "epoch": 1.8846202773184073, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.090651512145996, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8719236850738525, "num_tokens": 565082114.0, "step": 14815 }, { "epoch": 1.8847474875969978, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.14290714263916, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8696252107620239, "num_tokens": 565118350.0, "step": 14816 }, { "epoch": 1.8848746978755884, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.167734146118164, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8626171350479126, "num_tokens": 565153313.0, "step": 14817 }, { "epoch": 1.885001908154179, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.058297157287598, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8576244711875916, "num_tokens": 565191332.0, "step": 14818 }, { "epoch": 1.8851291184327694, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.105192184448242, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8663256168365479, "num_tokens": 565229937.0, "step": 14819 }, { "epoch": 1.88525632871136, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.064849853515625, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8648699522018433, "num_tokens": 565266271.0, "step": 14820 }, { "epoch": 1.8853835389899505, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.107810020446777, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8676432967185974, "num_tokens": 565300593.0, "step": 14821 }, { "epoch": 1.885510749268541, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.13232135772705, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8689461946487427, "num_tokens": 565341801.0, "step": 14822 }, { "epoch": 1.8856379595471315, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.110689163208008, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8871341943740845, "num_tokens": 565374380.0, "step": 14823 }, { "epoch": 1.885765169825722, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.087916374206543, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8619286417961121, "num_tokens": 565413096.0, "step": 14824 }, { "epoch": 1.8858923801043126, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.14428997039795, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8795763254165649, "num_tokens": 565453292.0, "step": 14825 }, { "epoch": 1.886019590382903, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.091551780700684, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8644425868988037, "num_tokens": 565489311.0, "step": 14826 }, { "epoch": 1.8861468006614934, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.1168794631958, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8536834716796875, "num_tokens": 565526293.0, "step": 14827 }, { "epoch": 1.886274010940084, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.092643737792969, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8632563948631287, "num_tokens": 565569215.0, "step": 14828 }, { "epoch": 1.8864012212186745, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.108041763305664, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8802958726882935, "num_tokens": 565610755.0, "step": 14829 }, { "epoch": 1.886528431497265, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.12697982788086, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8750042915344238, "num_tokens": 565641964.0, "step": 14830 }, { "epoch": 1.8866556417758555, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.1647310256958, "learning_rate": 1e-06, "loss": 0.511, "mean_token_accuracy": 0.8365319967269897, "num_tokens": 565682852.0, "step": 14831 }, { "epoch": 1.8867828520544458, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.057456016540527, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.886465311050415, "num_tokens": 565721352.0, "step": 14832 }, { "epoch": 1.8869100623330364, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.090461730957031, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8746776580810547, "num_tokens": 565764045.0, "step": 14833 }, { "epoch": 1.887037272611627, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.129127502441406, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8776252865791321, "num_tokens": 565797617.0, "step": 14834 }, { "epoch": 1.8871644828902174, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.142189025878906, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8685195446014404, "num_tokens": 565839254.0, "step": 14835 }, { "epoch": 1.887291693168808, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.155701637268066, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8699657917022705, "num_tokens": 565874742.0, "step": 14836 }, { "epoch": 1.8874189034473985, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.117227554321289, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.870344877243042, "num_tokens": 565911043.0, "step": 14837 }, { "epoch": 1.887546113725989, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.080426216125488, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8653777241706848, "num_tokens": 565950228.0, "step": 14838 }, { "epoch": 1.8876733240045795, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.147326469421387, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8818241953849792, "num_tokens": 565982891.0, "step": 14839 }, { "epoch": 1.88780053428317, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.043636322021484, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8685349822044373, "num_tokens": 566022601.0, "step": 14840 }, { "epoch": 1.8879277445617606, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.135891914367676, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.870649516582489, "num_tokens": 566054928.0, "step": 14841 }, { "epoch": 1.8880549548403511, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.083908081054688, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8637727499008179, "num_tokens": 566094900.0, "step": 14842 }, { "epoch": 1.8881821651189417, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.127142906188965, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8675997257232666, "num_tokens": 566125230.0, "step": 14843 }, { "epoch": 1.8883093753975322, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.086055755615234, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8846775889396667, "num_tokens": 566165320.0, "step": 14844 }, { "epoch": 1.8884365856761227, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.060918807983398, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.859882116317749, "num_tokens": 566202769.0, "step": 14845 }, { "epoch": 1.8885637959547132, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.091825485229492, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8716107606887817, "num_tokens": 566245874.0, "step": 14846 }, { "epoch": 1.8886910062333038, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.091822624206543, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8517003059387207, "num_tokens": 566284855.0, "step": 14847 }, { "epoch": 1.8888182165118943, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.086434364318848, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8623000979423523, "num_tokens": 566323422.0, "step": 14848 }, { "epoch": 1.8889454267904848, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.158336639404297, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8778638243675232, "num_tokens": 566364357.0, "step": 14849 }, { "epoch": 1.8890726370690751, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.063966751098633, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8670473694801331, "num_tokens": 566404498.0, "step": 14850 }, { "epoch": 1.8891998473476657, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.168712615966797, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8660987615585327, "num_tokens": 566442960.0, "step": 14851 }, { "epoch": 1.8893270576262562, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.152950286865234, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8782011866569519, "num_tokens": 566475087.0, "step": 14852 }, { "epoch": 1.8894542679048467, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.089022636413574, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8604819774627686, "num_tokens": 566514337.0, "step": 14853 }, { "epoch": 1.8895814781834372, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.122566223144531, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8603003025054932, "num_tokens": 566553204.0, "step": 14854 }, { "epoch": 1.8897086884620278, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.11506175994873, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8672199249267578, "num_tokens": 566589697.0, "step": 14855 }, { "epoch": 1.889835898740618, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.053908348083496, "learning_rate": 1e-06, "loss": 0.4556, "mean_token_accuracy": 0.8552770614624023, "num_tokens": 566629680.0, "step": 14856 }, { "epoch": 1.8899631090192086, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.060754776000977, "learning_rate": 1e-06, "loss": 0.334, "mean_token_accuracy": 0.891830563545227, "num_tokens": 566670931.0, "step": 14857 }, { "epoch": 1.8900903192977991, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.109009742736816, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8645387887954712, "num_tokens": 566706673.0, "step": 14858 }, { "epoch": 1.8902175295763897, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.102517127990723, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8655661344528198, "num_tokens": 566749503.0, "step": 14859 }, { "epoch": 1.8903447398549802, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.096779823303223, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8836696147918701, "num_tokens": 566790700.0, "step": 14860 }, { "epoch": 1.8904719501335707, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.039566040039062, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8686871528625488, "num_tokens": 566832995.0, "step": 14861 }, { "epoch": 1.8905991604121613, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.131930351257324, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8572974801063538, "num_tokens": 566868190.0, "step": 14862 }, { "epoch": 1.8907263706907518, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.111330032348633, "learning_rate": 1e-06, "loss": 0.4916, "mean_token_accuracy": 0.8410544395446777, "num_tokens": 566906637.0, "step": 14863 }, { "epoch": 1.8908535809693423, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.111662864685059, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8689411878585815, "num_tokens": 566946006.0, "step": 14864 }, { "epoch": 1.8909807912479328, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.088104248046875, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8574090003967285, "num_tokens": 566987726.0, "step": 14865 }, { "epoch": 1.8911080015265234, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.10378646850586, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.884614109992981, "num_tokens": 567027432.0, "step": 14866 }, { "epoch": 1.891235211805114, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.101522445678711, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8749141097068787, "num_tokens": 567060607.0, "step": 14867 }, { "epoch": 1.8913624220837044, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.095952033996582, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8630006313323975, "num_tokens": 567100691.0, "step": 14868 }, { "epoch": 1.891489632362295, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.079039573669434, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.887687087059021, "num_tokens": 567139460.0, "step": 14869 }, { "epoch": 1.8916168426408855, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.075246810913086, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8577343821525574, "num_tokens": 567185384.0, "step": 14870 }, { "epoch": 1.891744052919476, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.127311706542969, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8687381744384766, "num_tokens": 567219898.0, "step": 14871 }, { "epoch": 1.8918712631980665, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.069302558898926, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8767800331115723, "num_tokens": 567260322.0, "step": 14872 }, { "epoch": 1.891998473476657, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.074185371398926, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8650162220001221, "num_tokens": 567293739.0, "step": 14873 }, { "epoch": 1.8921256837552476, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.139132499694824, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8709630966186523, "num_tokens": 567327282.0, "step": 14874 }, { "epoch": 1.892252894033838, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.071027755737305, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8632412552833557, "num_tokens": 567366402.0, "step": 14875 }, { "epoch": 1.8923801043124284, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.192255973815918, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.862651526927948, "num_tokens": 567409924.0, "step": 14876 }, { "epoch": 1.892507314591019, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.107450485229492, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8725345134735107, "num_tokens": 567450355.0, "step": 14877 }, { "epoch": 1.8926345248696095, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.096875190734863, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8826977610588074, "num_tokens": 567487701.0, "step": 14878 }, { "epoch": 1.8927617351482, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.10014533996582, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8686962127685547, "num_tokens": 567527619.0, "step": 14879 }, { "epoch": 1.8928889454267905, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.09882926940918, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8686752319335938, "num_tokens": 567567016.0, "step": 14880 }, { "epoch": 1.8930161557053808, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.153419494628906, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8566321134567261, "num_tokens": 567605465.0, "step": 14881 }, { "epoch": 1.8931433659839714, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.102055549621582, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.872395396232605, "num_tokens": 567646680.0, "step": 14882 }, { "epoch": 1.893270576262562, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.126445770263672, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8702342510223389, "num_tokens": 567685685.0, "step": 14883 }, { "epoch": 1.8933977865411524, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.149301528930664, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8585333228111267, "num_tokens": 567717222.0, "step": 14884 }, { "epoch": 1.893524996819743, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.086029052734375, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8594128489494324, "num_tokens": 567761050.0, "step": 14885 }, { "epoch": 1.8936522070983335, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.123945236206055, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8677898645401001, "num_tokens": 567801423.0, "step": 14886 }, { "epoch": 1.893779417376924, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.129528045654297, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8598053455352783, "num_tokens": 567844614.0, "step": 14887 }, { "epoch": 1.8939066276555145, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.150970458984375, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8716880083084106, "num_tokens": 567878803.0, "step": 14888 }, { "epoch": 1.894033837934105, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.112564086914062, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.881251335144043, "num_tokens": 567909419.0, "step": 14889 }, { "epoch": 1.8941610482126956, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.149393081665039, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8703312873840332, "num_tokens": 567949221.0, "step": 14890 }, { "epoch": 1.8942882584912861, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.132905006408691, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8853743076324463, "num_tokens": 567985101.0, "step": 14891 }, { "epoch": 1.8944154687698767, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.09981632232666, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8521294593811035, "num_tokens": 568030542.0, "step": 14892 }, { "epoch": 1.8945426790484672, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.153668403625488, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8719079494476318, "num_tokens": 568070723.0, "step": 14893 }, { "epoch": 1.8946698893270577, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.132702827453613, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.863545298576355, "num_tokens": 568113444.0, "step": 14894 }, { "epoch": 1.8947970996056482, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.089703559875488, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8733578324317932, "num_tokens": 568152646.0, "step": 14895 }, { "epoch": 1.8949243098842388, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.145160675048828, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8691164255142212, "num_tokens": 568187046.0, "step": 14896 }, { "epoch": 1.8950515201628293, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.179842948913574, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8647047281265259, "num_tokens": 568225052.0, "step": 14897 }, { "epoch": 1.8951787304414198, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.095956802368164, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8619539737701416, "num_tokens": 568258299.0, "step": 14898 }, { "epoch": 1.8953059407200101, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.159997940063477, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8752843737602234, "num_tokens": 568297137.0, "step": 14899 }, { "epoch": 1.8954331509986007, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.057514190673828, "learning_rate": 1e-06, "loss": 0.3388, "mean_token_accuracy": 0.8908596038818359, "num_tokens": 568331102.0, "step": 14900 }, { "epoch": 1.8955603612771912, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.14036750793457, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8671072721481323, "num_tokens": 568373926.0, "step": 14901 }, { "epoch": 1.8956875715557817, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.143410682678223, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8786441087722778, "num_tokens": 568406470.0, "step": 14902 }, { "epoch": 1.8958147818343722, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.102066040039062, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8759867548942566, "num_tokens": 568443452.0, "step": 14903 }, { "epoch": 1.8959419921129628, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.142788887023926, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8647701740264893, "num_tokens": 568480085.0, "step": 14904 }, { "epoch": 1.896069202391553, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.126007080078125, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8667951822280884, "num_tokens": 568519191.0, "step": 14905 }, { "epoch": 1.8961964126701436, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.093926429748535, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8755344152450562, "num_tokens": 568554224.0, "step": 14906 }, { "epoch": 1.8963236229487341, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.110386848449707, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8862932920455933, "num_tokens": 568586102.0, "step": 14907 }, { "epoch": 1.8964508332273247, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.16068172454834, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.864082396030426, "num_tokens": 568625224.0, "step": 14908 }, { "epoch": 1.8965780435059152, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.107648849487305, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.8484233617782593, "num_tokens": 568665588.0, "step": 14909 }, { "epoch": 1.8967052537845057, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.142951965332031, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8857553005218506, "num_tokens": 568698408.0, "step": 14910 }, { "epoch": 1.8968324640630962, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.10279655456543, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8573408722877502, "num_tokens": 568731982.0, "step": 14911 }, { "epoch": 1.8969596743416868, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.18315601348877, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.868841290473938, "num_tokens": 568769388.0, "step": 14912 }, { "epoch": 1.8970868846202773, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.107102394104004, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8759251236915588, "num_tokens": 568810329.0, "step": 14913 }, { "epoch": 1.8972140948988678, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.088000297546387, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8744043707847595, "num_tokens": 568852241.0, "step": 14914 }, { "epoch": 1.8973413051774584, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.137182235717773, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8643912076950073, "num_tokens": 568889660.0, "step": 14915 }, { "epoch": 1.8974685154560489, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.117358207702637, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8669297695159912, "num_tokens": 568934489.0, "step": 14916 }, { "epoch": 1.8975957257346394, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.096148490905762, "learning_rate": 1e-06, "loss": 0.3364, "mean_token_accuracy": 0.8913748264312744, "num_tokens": 568971904.0, "step": 14917 }, { "epoch": 1.89772293601323, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.066225051879883, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8507500886917114, "num_tokens": 569011863.0, "step": 14918 }, { "epoch": 1.8978501462918205, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.102676391601562, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8909135460853577, "num_tokens": 569050295.0, "step": 14919 }, { "epoch": 1.897977356570411, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.132648468017578, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8629285097122192, "num_tokens": 569087013.0, "step": 14920 }, { "epoch": 1.8981045668490015, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.103083610534668, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8632510900497437, "num_tokens": 569125226.0, "step": 14921 }, { "epoch": 1.898231777127592, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.098021507263184, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8661530017852783, "num_tokens": 569161910.0, "step": 14922 }, { "epoch": 1.8983589874061826, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.12458324432373, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8725610971450806, "num_tokens": 569196519.0, "step": 14923 }, { "epoch": 1.898486197684773, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.141023635864258, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8653698563575745, "num_tokens": 569234404.0, "step": 14924 }, { "epoch": 1.8986134079633634, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.115179061889648, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8607239723205566, "num_tokens": 569272728.0, "step": 14925 }, { "epoch": 1.898740618241954, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.129960060119629, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8619715571403503, "num_tokens": 569315434.0, "step": 14926 }, { "epoch": 1.8988678285205445, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.102666854858398, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8527787327766418, "num_tokens": 569357473.0, "step": 14927 }, { "epoch": 1.898995038799135, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.129829406738281, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.8576629757881165, "num_tokens": 569398747.0, "step": 14928 }, { "epoch": 1.8991222490777255, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.14027214050293, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8763765096664429, "num_tokens": 569434855.0, "step": 14929 }, { "epoch": 1.8992494593563158, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.085661888122559, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8683671951293945, "num_tokens": 569475236.0, "step": 14930 }, { "epoch": 1.8993766696349064, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.146855354309082, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8678746223449707, "num_tokens": 569513599.0, "step": 14931 }, { "epoch": 1.899503879913497, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.160079002380371, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8757350444793701, "num_tokens": 569559059.0, "step": 14932 }, { "epoch": 1.8996310901920874, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.130937576293945, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8865358829498291, "num_tokens": 569595240.0, "step": 14933 }, { "epoch": 1.899758300470678, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.148344039916992, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8604377508163452, "num_tokens": 569629336.0, "step": 14934 }, { "epoch": 1.8998855107492685, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.099505424499512, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8803722858428955, "num_tokens": 569663613.0, "step": 14935 }, { "epoch": 1.900012721027859, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.142190933227539, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8541210889816284, "num_tokens": 569702884.0, "step": 14936 }, { "epoch": 1.9001399313064495, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.123966217041016, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.884305477142334, "num_tokens": 569744618.0, "step": 14937 }, { "epoch": 1.90026714158504, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.16468334197998, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8787316679954529, "num_tokens": 569777083.0, "step": 14938 }, { "epoch": 1.9003943518636306, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.12674331665039, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8762936592102051, "num_tokens": 569809652.0, "step": 14939 }, { "epoch": 1.9005215621422211, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.139738082885742, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8741791844367981, "num_tokens": 569846465.0, "step": 14940 }, { "epoch": 1.9006487724208116, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.169429779052734, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8630890846252441, "num_tokens": 569883655.0, "step": 14941 }, { "epoch": 1.9007759826994022, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.1233549118042, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8677790760993958, "num_tokens": 569923358.0, "step": 14942 }, { "epoch": 1.9009031929779927, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.150389671325684, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8663254976272583, "num_tokens": 569961393.0, "step": 14943 }, { "epoch": 1.9010304032565832, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.15445327758789, "learning_rate": 1e-06, "loss": 0.4955, "mean_token_accuracy": 0.8423005938529968, "num_tokens": 570002092.0, "step": 14944 }, { "epoch": 1.9011576135351738, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.100964546203613, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8549315929412842, "num_tokens": 570037699.0, "step": 14945 }, { "epoch": 1.9012848238137643, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.093385696411133, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8473814129829407, "num_tokens": 570077693.0, "step": 14946 }, { "epoch": 1.9014120340923548, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.134932518005371, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8886861205101013, "num_tokens": 570115354.0, "step": 14947 }, { "epoch": 1.9015392443709451, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.087934494018555, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8776624202728271, "num_tokens": 570154387.0, "step": 14948 }, { "epoch": 1.9016664546495357, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.229575157165527, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8647428750991821, "num_tokens": 570189743.0, "step": 14949 }, { "epoch": 1.9017936649281262, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.139212608337402, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8812214136123657, "num_tokens": 570224009.0, "step": 14950 }, { "epoch": 1.9019208752067167, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.09700870513916, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8759223222732544, "num_tokens": 570266703.0, "step": 14951 }, { "epoch": 1.9020480854853072, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.177191734313965, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8686949014663696, "num_tokens": 570308241.0, "step": 14952 }, { "epoch": 1.9021752957638978, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.09547233581543, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8690186738967896, "num_tokens": 570346931.0, "step": 14953 }, { "epoch": 1.902302506042488, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.081191062927246, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8827964067459106, "num_tokens": 570380373.0, "step": 14954 }, { "epoch": 1.9024297163210786, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.081924438476562, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8804901838302612, "num_tokens": 570420106.0, "step": 14955 }, { "epoch": 1.9025569265996691, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.10017204284668, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.8898226022720337, "num_tokens": 570464534.0, "step": 14956 }, { "epoch": 1.9026841368782597, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.083264350891113, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8594028949737549, "num_tokens": 570505734.0, "step": 14957 }, { "epoch": 1.9028113471568502, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.05345630645752, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8784376382827759, "num_tokens": 570545913.0, "step": 14958 }, { "epoch": 1.9029385574354407, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.154154777526855, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8766965270042419, "num_tokens": 570584397.0, "step": 14959 }, { "epoch": 1.9030657677140312, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.09809398651123, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8755195140838623, "num_tokens": 570622285.0, "step": 14960 }, { "epoch": 1.9031929779926218, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.147493362426758, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8542678952217102, "num_tokens": 570656331.0, "step": 14961 }, { "epoch": 1.9033201882712123, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.10063362121582, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8736343383789062, "num_tokens": 570699944.0, "step": 14962 }, { "epoch": 1.9034473985498028, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.167157173156738, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8637498617172241, "num_tokens": 570737649.0, "step": 14963 }, { "epoch": 1.9035746088283934, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.139404296875, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8667929172515869, "num_tokens": 570773193.0, "step": 14964 }, { "epoch": 1.9037018191069839, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.175931930541992, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8638918995857239, "num_tokens": 570809117.0, "step": 14965 }, { "epoch": 1.9038290293855744, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.134360313415527, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8814130425453186, "num_tokens": 570844268.0, "step": 14966 }, { "epoch": 1.903956239664165, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.17093276977539, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8613986968994141, "num_tokens": 570890166.0, "step": 14967 }, { "epoch": 1.9040834499427555, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.117443084716797, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8651471138000488, "num_tokens": 570921419.0, "step": 14968 }, { "epoch": 1.904210660221346, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.157679557800293, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8600344657897949, "num_tokens": 570966444.0, "step": 14969 }, { "epoch": 1.9043378704999365, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.069398880004883, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8794552087783813, "num_tokens": 571011656.0, "step": 14970 }, { "epoch": 1.904465080778527, "ewc_loss": 0.03271484375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 13.11701774597168, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8737844228744507, "num_tokens": 571047828.0, "step": 14971 }, { "epoch": 1.9045922910571176, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.095674514770508, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8677552342414856, "num_tokens": 571084541.0, "step": 14972 }, { "epoch": 1.9047195013357079, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.191871643066406, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8634988069534302, "num_tokens": 571120978.0, "step": 14973 }, { "epoch": 1.9048467116142984, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.094101905822754, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8837291598320007, "num_tokens": 571160181.0, "step": 14974 }, { "epoch": 1.904973921892889, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.136743545532227, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8755035400390625, "num_tokens": 571194648.0, "step": 14975 }, { "epoch": 1.9051011321714795, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.084458351135254, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8654884099960327, "num_tokens": 571238449.0, "step": 14976 }, { "epoch": 1.90522834245007, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.189153671264648, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8705861568450928, "num_tokens": 571277266.0, "step": 14977 }, { "epoch": 1.9053555527286605, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.142333984375, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8806119561195374, "num_tokens": 571316891.0, "step": 14978 }, { "epoch": 1.9054827630072508, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.108296394348145, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8720386028289795, "num_tokens": 571353097.0, "step": 14979 }, { "epoch": 1.9056099732858414, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.133523941040039, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8839256167411804, "num_tokens": 571392217.0, "step": 14980 }, { "epoch": 1.905737183564432, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.092989921569824, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8779768943786621, "num_tokens": 571430440.0, "step": 14981 }, { "epoch": 1.9058643938430224, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.155153274536133, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8660783767700195, "num_tokens": 571472002.0, "step": 14982 }, { "epoch": 1.905991604121613, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.139833450317383, "learning_rate": 1e-06, "loss": 0.4738, "mean_token_accuracy": 0.8508181571960449, "num_tokens": 571509368.0, "step": 14983 }, { "epoch": 1.9061188144002035, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.138311386108398, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.855482816696167, "num_tokens": 571543898.0, "step": 14984 }, { "epoch": 1.906246024678794, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.07666015625, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8643470406532288, "num_tokens": 571584395.0, "step": 14985 }, { "epoch": 1.9063732349573845, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.106985092163086, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8685615062713623, "num_tokens": 571624792.0, "step": 14986 }, { "epoch": 1.906500445235975, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.128924369812012, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8713529109954834, "num_tokens": 571667197.0, "step": 14987 }, { "epoch": 1.9066276555145656, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.114177703857422, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8641458749771118, "num_tokens": 571703648.0, "step": 14988 }, { "epoch": 1.9067548657931561, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.185691833496094, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8618307113647461, "num_tokens": 571738980.0, "step": 14989 }, { "epoch": 1.9068820760717466, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.143256187438965, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8718627691268921, "num_tokens": 571774540.0, "step": 14990 }, { "epoch": 1.9070092863503372, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.200862884521484, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8748908638954163, "num_tokens": 571805286.0, "step": 14991 }, { "epoch": 1.9071364966289277, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.137770652770996, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8636949062347412, "num_tokens": 571847936.0, "step": 14992 }, { "epoch": 1.9072637069075182, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.135289192199707, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8656982183456421, "num_tokens": 571885454.0, "step": 14993 }, { "epoch": 1.9073909171861088, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.13273811340332, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8681902289390564, "num_tokens": 571924870.0, "step": 14994 }, { "epoch": 1.9075181274646993, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.175217628479004, "learning_rate": 1e-06, "loss": 0.4825, "mean_token_accuracy": 0.8492200374603271, "num_tokens": 571957870.0, "step": 14995 }, { "epoch": 1.9076453377432898, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.165238380432129, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8618403673171997, "num_tokens": 571993645.0, "step": 14996 }, { "epoch": 1.9077725480218801, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.138665199279785, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8782042264938354, "num_tokens": 572028594.0, "step": 14997 }, { "epoch": 1.9078997583004706, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.126108169555664, "learning_rate": 1e-06, "loss": 0.4903, "mean_token_accuracy": 0.8437784314155579, "num_tokens": 572066971.0, "step": 14998 }, { "epoch": 1.9080269685790612, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.071786880493164, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8735014796257019, "num_tokens": 572107609.0, "step": 14999 }, { "epoch": 1.9081541788576517, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.123274803161621, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8785566091537476, "num_tokens": 572146940.0, "step": 15000 }, { "epoch": 1.9082813891362422, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.134064674377441, "learning_rate": 1e-06, "loss": 0.3326, "mean_token_accuracy": 0.8955695033073425, "num_tokens": 572184010.0, "step": 15001 }, { "epoch": 1.9084085994148328, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.101099967956543, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.853574812412262, "num_tokens": 572223184.0, "step": 15002 }, { "epoch": 1.908535809693423, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.14687442779541, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8696092367172241, "num_tokens": 572261770.0, "step": 15003 }, { "epoch": 1.9086630199720136, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.11374282836914, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8717187643051147, "num_tokens": 572303229.0, "step": 15004 }, { "epoch": 1.9087902302506041, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.18586254119873, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8579221963882446, "num_tokens": 572337839.0, "step": 15005 }, { "epoch": 1.9089174405291947, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.09328842163086, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8687834143638611, "num_tokens": 572373604.0, "step": 15006 }, { "epoch": 1.9090446508077852, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.139713287353516, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8767929077148438, "num_tokens": 572410425.0, "step": 15007 }, { "epoch": 1.9091718610863757, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.0994291305542, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8786431550979614, "num_tokens": 572443998.0, "step": 15008 }, { "epoch": 1.9092990713649662, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.169727325439453, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8571830987930298, "num_tokens": 572485337.0, "step": 15009 }, { "epoch": 1.9094262816435568, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.126569747924805, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8591758608818054, "num_tokens": 572528273.0, "step": 15010 }, { "epoch": 1.9095534919221473, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.148740768432617, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8723580241203308, "num_tokens": 572569467.0, "step": 15011 }, { "epoch": 1.9096807022007378, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.056254386901855, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8770825862884521, "num_tokens": 572609626.0, "step": 15012 }, { "epoch": 1.9098079124793284, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.134347915649414, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8535935878753662, "num_tokens": 572652907.0, "step": 15013 }, { "epoch": 1.9099351227579189, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.168009757995605, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8742047548294067, "num_tokens": 572687460.0, "step": 15014 }, { "epoch": 1.9100623330365094, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.148490905761719, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8717004060745239, "num_tokens": 572723937.0, "step": 15015 }, { "epoch": 1.9101895433151, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.07624626159668, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8650244474411011, "num_tokens": 572764929.0, "step": 15016 }, { "epoch": 1.9103167535936905, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 13.124492645263672, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8670938611030579, "num_tokens": 572802545.0, "step": 15017 }, { "epoch": 1.910443963872281, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.075943946838379, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8639484643936157, "num_tokens": 572850008.0, "step": 15018 }, { "epoch": 1.9105711741508715, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.110864639282227, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8774796724319458, "num_tokens": 572888968.0, "step": 15019 }, { "epoch": 1.910698384429462, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.137922286987305, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8683469295501709, "num_tokens": 572926314.0, "step": 15020 }, { "epoch": 1.9108255947080524, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.143383979797363, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8736498355865479, "num_tokens": 572961789.0, "step": 15021 }, { "epoch": 1.9109528049866429, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.17432975769043, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8712206482887268, "num_tokens": 572993680.0, "step": 15022 }, { "epoch": 1.9110800152652334, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.163825035095215, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8707902431488037, "num_tokens": 573024003.0, "step": 15023 }, { "epoch": 1.911207225543824, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.13847827911377, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8771917223930359, "num_tokens": 573055138.0, "step": 15024 }, { "epoch": 1.9113344358224145, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.111825942993164, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8742974996566772, "num_tokens": 573092373.0, "step": 15025 }, { "epoch": 1.911461646101005, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.183629989624023, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8647288084030151, "num_tokens": 573131486.0, "step": 15026 }, { "epoch": 1.9115888563795955, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.125048637390137, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8787726759910583, "num_tokens": 573170114.0, "step": 15027 }, { "epoch": 1.9117160666581858, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.105709075927734, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8729939460754395, "num_tokens": 573200906.0, "step": 15028 }, { "epoch": 1.9118432769367764, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.115754127502441, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8567456603050232, "num_tokens": 573243731.0, "step": 15029 }, { "epoch": 1.9119704872153669, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.126546859741211, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8758245706558228, "num_tokens": 573283335.0, "step": 15030 }, { "epoch": 1.9120976974939574, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.134069442749023, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8571351766586304, "num_tokens": 573326628.0, "step": 15031 }, { "epoch": 1.912224907772548, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.087491989135742, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8753596544265747, "num_tokens": 573369025.0, "step": 15032 }, { "epoch": 1.9123521180511385, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.125354766845703, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8630349636077881, "num_tokens": 573411923.0, "step": 15033 }, { "epoch": 1.912479328329729, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.18642807006836, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8653416633605957, "num_tokens": 573450455.0, "step": 15034 }, { "epoch": 1.9126065386083195, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.089909553527832, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8604294657707214, "num_tokens": 573492886.0, "step": 15035 }, { "epoch": 1.91273374888691, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.1351957321167, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8749903440475464, "num_tokens": 573532073.0, "step": 15036 }, { "epoch": 1.9128609591655006, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.084229469299316, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.885382354259491, "num_tokens": 573573632.0, "step": 15037 }, { "epoch": 1.9129881694440911, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.112317085266113, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8795627355575562, "num_tokens": 573610194.0, "step": 15038 }, { "epoch": 1.9131153797226816, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.101741790771484, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8618164658546448, "num_tokens": 573650237.0, "step": 15039 }, { "epoch": 1.9132425900012722, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.106537818908691, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8732463121414185, "num_tokens": 573686859.0, "step": 15040 }, { "epoch": 1.9133698002798627, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.137478828430176, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8699997663497925, "num_tokens": 573726770.0, "step": 15041 }, { "epoch": 1.9134970105584532, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.09509563446045, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8640345335006714, "num_tokens": 573768668.0, "step": 15042 }, { "epoch": 1.9136242208370438, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.158429145812988, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8604704141616821, "num_tokens": 573807880.0, "step": 15043 }, { "epoch": 1.9137514311156343, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.101040840148926, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8675386905670166, "num_tokens": 573847018.0, "step": 15044 }, { "epoch": 1.9138786413942248, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.072916984558105, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8850897550582886, "num_tokens": 573881193.0, "step": 15045 }, { "epoch": 1.9140058516728151, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.150615692138672, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8785852193832397, "num_tokens": 573918457.0, "step": 15046 }, { "epoch": 1.9141330619514056, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.155370712280273, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8884677886962891, "num_tokens": 573958818.0, "step": 15047 }, { "epoch": 1.9142602722299962, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.116183280944824, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8642977476119995, "num_tokens": 573996735.0, "step": 15048 }, { "epoch": 1.9143874825085867, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.14111042022705, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8735926151275635, "num_tokens": 574035005.0, "step": 15049 }, { "epoch": 1.9145146927871772, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.132399559020996, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8686622381210327, "num_tokens": 574077817.0, "step": 15050 }, { "epoch": 1.9146419030657678, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.11849594116211, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8774466514587402, "num_tokens": 574113083.0, "step": 15051 }, { "epoch": 1.914769113344358, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.13990592956543, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8739847540855408, "num_tokens": 574147324.0, "step": 15052 }, { "epoch": 1.9148963236229486, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.084315299987793, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8849283456802368, "num_tokens": 574186108.0, "step": 15053 }, { "epoch": 1.9150235339015391, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.18057918548584, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8826661705970764, "num_tokens": 574218076.0, "step": 15054 }, { "epoch": 1.9151507441801296, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.106340408325195, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8802503347396851, "num_tokens": 574256416.0, "step": 15055 }, { "epoch": 1.9152779544587202, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.103843688964844, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8833638429641724, "num_tokens": 574295262.0, "step": 15056 }, { "epoch": 1.9154051647373107, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.1624174118042, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8595608472824097, "num_tokens": 574329836.0, "step": 15057 }, { "epoch": 1.9155323750159012, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.126852035522461, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8733401298522949, "num_tokens": 574363189.0, "step": 15058 }, { "epoch": 1.9156595852944918, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.138904571533203, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8781248927116394, "num_tokens": 574396374.0, "step": 15059 }, { "epoch": 1.9157867955730823, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.158690452575684, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.872908353805542, "num_tokens": 574435488.0, "step": 15060 }, { "epoch": 1.9159140058516728, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.161458969116211, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8704670667648315, "num_tokens": 574475261.0, "step": 15061 }, { "epoch": 1.9160412161302633, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.151447296142578, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.8890591263771057, "num_tokens": 574513531.0, "step": 15062 }, { "epoch": 1.9161684264088539, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.175490379333496, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8555202484130859, "num_tokens": 574551736.0, "step": 15063 }, { "epoch": 1.9162956366874444, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.172078132629395, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8643986582756042, "num_tokens": 574589978.0, "step": 15064 }, { "epoch": 1.916422846966035, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.185868263244629, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8492578268051147, "num_tokens": 574628845.0, "step": 15065 }, { "epoch": 1.9165500572446255, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.163710594177246, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8669041991233826, "num_tokens": 574660700.0, "step": 15066 }, { "epoch": 1.916677267523216, "ewc_loss": 0.032958984375, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.131580352783203, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8621464967727661, "num_tokens": 574692195.0, "step": 15067 }, { "epoch": 1.9168044778018065, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.128474235534668, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8607272505760193, "num_tokens": 574732462.0, "step": 15068 }, { "epoch": 1.916931688080397, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.101029396057129, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8772204518318176, "num_tokens": 574769353.0, "step": 15069 }, { "epoch": 1.9170588983589874, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.173726081848145, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.870246171951294, "num_tokens": 574804958.0, "step": 15070 }, { "epoch": 1.9171861086375779, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.105022430419922, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8699862957000732, "num_tokens": 574843966.0, "step": 15071 }, { "epoch": 1.9173133189161684, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.089780807495117, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8744558691978455, "num_tokens": 574882916.0, "step": 15072 }, { "epoch": 1.917440529194759, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.14098072052002, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8787869215011597, "num_tokens": 574922154.0, "step": 15073 }, { "epoch": 1.9175677394733495, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.163402557373047, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8673720359802246, "num_tokens": 574956894.0, "step": 15074 }, { "epoch": 1.91769494975194, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.142274856567383, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8848588466644287, "num_tokens": 574998343.0, "step": 15075 }, { "epoch": 1.9178221600305305, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.161402702331543, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8794572949409485, "num_tokens": 575036689.0, "step": 15076 }, { "epoch": 1.9179493703091208, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.150715827941895, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8643493056297302, "num_tokens": 575075510.0, "step": 15077 }, { "epoch": 1.9180765805877114, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.124899864196777, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8680790662765503, "num_tokens": 575115797.0, "step": 15078 }, { "epoch": 1.9182037908663019, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.2055082321167, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8570282459259033, "num_tokens": 575155240.0, "step": 15079 }, { "epoch": 1.9183310011448924, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.192012786865234, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8573102355003357, "num_tokens": 575191048.0, "step": 15080 }, { "epoch": 1.918458211423483, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.203683853149414, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8591873049736023, "num_tokens": 575227778.0, "step": 15081 }, { "epoch": 1.9185854217020735, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.080398559570312, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8712284564971924, "num_tokens": 575269774.0, "step": 15082 }, { "epoch": 1.918712631980664, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.18115234375, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8616655468940735, "num_tokens": 575310710.0, "step": 15083 }, { "epoch": 1.9188398422592545, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.153907775878906, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8779086470603943, "num_tokens": 575344073.0, "step": 15084 }, { "epoch": 1.918967052537845, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.152892112731934, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8612004518508911, "num_tokens": 575386737.0, "step": 15085 }, { "epoch": 1.9190942628164356, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.166335105895996, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8738738298416138, "num_tokens": 575424287.0, "step": 15086 }, { "epoch": 1.919221473095026, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.163069725036621, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8720055222511292, "num_tokens": 575463859.0, "step": 15087 }, { "epoch": 1.9193486833736166, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.185254096984863, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.865117609500885, "num_tokens": 575500838.0, "step": 15088 }, { "epoch": 1.9194758936522072, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.19578742980957, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8614148497581482, "num_tokens": 575535164.0, "step": 15089 }, { "epoch": 1.9196031039307977, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.124205589294434, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8766686916351318, "num_tokens": 575569493.0, "step": 15090 }, { "epoch": 1.9197303142093882, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.199134826660156, "learning_rate": 1e-06, "loss": 0.4672, "mean_token_accuracy": 0.8549558520317078, "num_tokens": 575603396.0, "step": 15091 }, { "epoch": 1.9198575244879788, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.177045822143555, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8766201734542847, "num_tokens": 575638255.0, "step": 15092 }, { "epoch": 1.9199847347665693, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.097334861755371, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8592283725738525, "num_tokens": 575685281.0, "step": 15093 }, { "epoch": 1.9201119450451598, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.16818618774414, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8764691948890686, "num_tokens": 575725030.0, "step": 15094 }, { "epoch": 1.9202391553237501, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.171487808227539, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8650192618370056, "num_tokens": 575764237.0, "step": 15095 }, { "epoch": 1.9203663656023406, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.181458473205566, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8676064610481262, "num_tokens": 575796732.0, "step": 15096 }, { "epoch": 1.9204935758809312, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.213913917541504, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8740746974945068, "num_tokens": 575833133.0, "step": 15097 }, { "epoch": 1.9206207861595217, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.076189041137695, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8655188083648682, "num_tokens": 575870920.0, "step": 15098 }, { "epoch": 1.9207479964381122, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.153325080871582, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8720980286598206, "num_tokens": 575909132.0, "step": 15099 }, { "epoch": 1.9208752067167028, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.154223442077637, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8702809810638428, "num_tokens": 575947979.0, "step": 15100 }, { "epoch": 1.921002416995293, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.152215003967285, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8592519760131836, "num_tokens": 575988888.0, "step": 15101 }, { "epoch": 1.9211296272738836, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.099993705749512, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8723753094673157, "num_tokens": 576028614.0, "step": 15102 }, { "epoch": 1.9212568375524741, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.102191925048828, "learning_rate": 1e-06, "loss": 0.326, "mean_token_accuracy": 0.8974317312240601, "num_tokens": 576066648.0, "step": 15103 }, { "epoch": 1.9213840478310646, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.156777381896973, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8680053949356079, "num_tokens": 576104045.0, "step": 15104 }, { "epoch": 1.9215112581096552, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.12093448638916, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8626131415367126, "num_tokens": 576145419.0, "step": 15105 }, { "epoch": 1.9216384683882457, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.109238624572754, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8664491176605225, "num_tokens": 576191410.0, "step": 15106 }, { "epoch": 1.9217656786668362, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.169321060180664, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8646010160446167, "num_tokens": 576226383.0, "step": 15107 }, { "epoch": 1.9218928889454268, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.222128868103027, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8619137406349182, "num_tokens": 576259958.0, "step": 15108 }, { "epoch": 1.9220200992240173, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.157336235046387, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8727107048034668, "num_tokens": 576299467.0, "step": 15109 }, { "epoch": 1.9221473095026078, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.191604614257812, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8630004525184631, "num_tokens": 576334137.0, "step": 15110 }, { "epoch": 1.9222745197811983, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.157005310058594, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8656983971595764, "num_tokens": 576379582.0, "step": 15111 }, { "epoch": 1.9224017300597889, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.201332092285156, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8835102915763855, "num_tokens": 576414590.0, "step": 15112 }, { "epoch": 1.9225289403383794, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.18350601196289, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8615152835845947, "num_tokens": 576445023.0, "step": 15113 }, { "epoch": 1.92265615061697, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.2113037109375, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8662599325180054, "num_tokens": 576477861.0, "step": 15114 }, { "epoch": 1.9227833608955605, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.103282928466797, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8660814166069031, "num_tokens": 576513121.0, "step": 15115 }, { "epoch": 1.922910571174151, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.162535667419434, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8754593729972839, "num_tokens": 576551039.0, "step": 15116 }, { "epoch": 1.9230377814527415, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.115942001342773, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8623329401016235, "num_tokens": 576590651.0, "step": 15117 }, { "epoch": 1.923164991731332, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.219289779663086, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8604332208633423, "num_tokens": 576629119.0, "step": 15118 }, { "epoch": 1.9232922020099223, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.170899391174316, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8783434629440308, "num_tokens": 576664240.0, "step": 15119 }, { "epoch": 1.9234194122885129, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.154864311218262, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8641047477722168, "num_tokens": 576701946.0, "step": 15120 }, { "epoch": 1.9235466225671034, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.233830451965332, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8806077837944031, "num_tokens": 576733541.0, "step": 15121 }, { "epoch": 1.923673832845694, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.18867015838623, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8560547828674316, "num_tokens": 576770930.0, "step": 15122 }, { "epoch": 1.9238010431242845, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.224993705749512, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8689686059951782, "num_tokens": 576807455.0, "step": 15123 }, { "epoch": 1.923928253402875, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.193719863891602, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8766588568687439, "num_tokens": 576849577.0, "step": 15124 }, { "epoch": 1.9240554636814655, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.185375213623047, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8633978366851807, "num_tokens": 576883506.0, "step": 15125 }, { "epoch": 1.9241826739600558, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.226346969604492, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8528668284416199, "num_tokens": 576923892.0, "step": 15126 }, { "epoch": 1.9243098842386464, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.184492111206055, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8867303729057312, "num_tokens": 576970832.0, "step": 15127 }, { "epoch": 1.9244370945172369, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.161052703857422, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8522791862487793, "num_tokens": 577006990.0, "step": 15128 }, { "epoch": 1.9245643047958274, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.15074348449707, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8756106495857239, "num_tokens": 577044932.0, "step": 15129 }, { "epoch": 1.924691515074418, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.219932556152344, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8530542254447937, "num_tokens": 577076217.0, "step": 15130 }, { "epoch": 1.9248187253530085, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.13296890258789, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8845013380050659, "num_tokens": 577112696.0, "step": 15131 }, { "epoch": 1.924945935631599, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.146624565124512, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8681635856628418, "num_tokens": 577156397.0, "step": 15132 }, { "epoch": 1.9250731459101895, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.233537673950195, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8729722499847412, "num_tokens": 577194192.0, "step": 15133 }, { "epoch": 1.92520035618878, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.199050903320312, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8646376132965088, "num_tokens": 577232639.0, "step": 15134 }, { "epoch": 1.9253275664673706, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.211963653564453, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.878432035446167, "num_tokens": 577267349.0, "step": 15135 }, { "epoch": 1.925454776745961, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.198641777038574, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8643911480903625, "num_tokens": 577308329.0, "step": 15136 }, { "epoch": 1.9255819870245516, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.134712219238281, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8708441257476807, "num_tokens": 577347825.0, "step": 15137 }, { "epoch": 1.9257091973031422, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.212911605834961, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8696926832199097, "num_tokens": 577379981.0, "step": 15138 }, { "epoch": 1.9258364075817327, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.168767929077148, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8727604746818542, "num_tokens": 577414134.0, "step": 15139 }, { "epoch": 1.9259636178603232, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.18240737915039, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8676409721374512, "num_tokens": 577451170.0, "step": 15140 }, { "epoch": 1.9260908281389137, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.239198684692383, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8563278913497925, "num_tokens": 577487447.0, "step": 15141 }, { "epoch": 1.9262180384175043, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.154074668884277, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8737388849258423, "num_tokens": 577525732.0, "step": 15142 }, { "epoch": 1.9263452486960948, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.165294647216797, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.86855149269104, "num_tokens": 577563303.0, "step": 15143 }, { "epoch": 1.926472458974685, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.131634712219238, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8831537961959839, "num_tokens": 577607592.0, "step": 15144 }, { "epoch": 1.9265996692532756, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.154661178588867, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8608642816543579, "num_tokens": 577651646.0, "step": 15145 }, { "epoch": 1.9267268795318662, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.19234848022461, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8773676753044128, "num_tokens": 577689530.0, "step": 15146 }, { "epoch": 1.9268540898104567, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.138710021972656, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8643639087677002, "num_tokens": 577732524.0, "step": 15147 }, { "epoch": 1.9269813000890472, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.188742637634277, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8743705153465271, "num_tokens": 577767164.0, "step": 15148 }, { "epoch": 1.9271085103676378, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.191798210144043, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.876718282699585, "num_tokens": 577802347.0, "step": 15149 }, { "epoch": 1.927235720646228, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.12998104095459, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8617250919342041, "num_tokens": 577842899.0, "step": 15150 }, { "epoch": 1.9273629309248186, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.203207969665527, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8690392971038818, "num_tokens": 577885137.0, "step": 15151 }, { "epoch": 1.9274901412034091, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.119512557983398, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8829214572906494, "num_tokens": 577925727.0, "step": 15152 }, { "epoch": 1.9276173514819996, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.188092231750488, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8727515935897827, "num_tokens": 577965399.0, "step": 15153 }, { "epoch": 1.9277445617605902, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.220376014709473, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8694772720336914, "num_tokens": 578007608.0, "step": 15154 }, { "epoch": 1.9278717720391807, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.154937744140625, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8807300329208374, "num_tokens": 578043425.0, "step": 15155 }, { "epoch": 1.9279989823177712, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.224964141845703, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8665816187858582, "num_tokens": 578075838.0, "step": 15156 }, { "epoch": 1.9281261925963618, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.202378273010254, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8636822700500488, "num_tokens": 578110113.0, "step": 15157 }, { "epoch": 1.9282534028749523, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.210290908813477, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8693329095840454, "num_tokens": 578145743.0, "step": 15158 }, { "epoch": 1.9283806131535428, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.173871994018555, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8775206804275513, "num_tokens": 578181588.0, "step": 15159 }, { "epoch": 1.9285078234321333, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.221212387084961, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8676840662956238, "num_tokens": 578222048.0, "step": 15160 }, { "epoch": 1.9286350337107239, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.143741607666016, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8722307682037354, "num_tokens": 578262664.0, "step": 15161 }, { "epoch": 1.9287622439893144, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.17773723602295, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8577660918235779, "num_tokens": 578303431.0, "step": 15162 }, { "epoch": 1.928889454267905, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.158597946166992, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8597084879875183, "num_tokens": 578338978.0, "step": 15163 }, { "epoch": 1.9290166645464955, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.194747924804688, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8651877045631409, "num_tokens": 578376340.0, "step": 15164 }, { "epoch": 1.929143874825086, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.162569046020508, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8714374303817749, "num_tokens": 578414714.0, "step": 15165 }, { "epoch": 1.9292710851036765, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.226656913757324, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.855556309223175, "num_tokens": 578453668.0, "step": 15166 }, { "epoch": 1.929398295382267, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.163113594055176, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8619341850280762, "num_tokens": 578494148.0, "step": 15167 }, { "epoch": 1.9295255056608573, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.199377059936523, "learning_rate": 1e-06, "loss": 0.4732, "mean_token_accuracy": 0.8496721982955933, "num_tokens": 578530743.0, "step": 15168 }, { "epoch": 1.9296527159394479, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.176657676696777, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8912179470062256, "num_tokens": 578567613.0, "step": 15169 }, { "epoch": 1.9297799262180384, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.16770076751709, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.867408275604248, "num_tokens": 578602981.0, "step": 15170 }, { "epoch": 1.929907136496629, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.176262855529785, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8632537126541138, "num_tokens": 578641027.0, "step": 15171 }, { "epoch": 1.9300343467752195, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 13.101117134094238, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8819206357002258, "num_tokens": 578683006.0, "step": 15172 }, { "epoch": 1.93016155705381, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.137941360473633, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8740687370300293, "num_tokens": 578716722.0, "step": 15173 }, { "epoch": 1.9302887673324005, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.169686317443848, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8500739932060242, "num_tokens": 578750740.0, "step": 15174 }, { "epoch": 1.9304159776109908, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.147305488586426, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8724303245544434, "num_tokens": 578791358.0, "step": 15175 }, { "epoch": 1.9305431878895813, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.155570030212402, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8538420796394348, "num_tokens": 578834514.0, "step": 15176 }, { "epoch": 1.9306703981681719, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.151281356811523, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8615204691886902, "num_tokens": 578875869.0, "step": 15177 }, { "epoch": 1.9307976084467624, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.172825813293457, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8733253479003906, "num_tokens": 578921079.0, "step": 15178 }, { "epoch": 1.930924818725353, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.123541831970215, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8686938285827637, "num_tokens": 578959837.0, "step": 15179 }, { "epoch": 1.9310520290039435, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.180801391601562, "learning_rate": 1e-06, "loss": 0.5107, "mean_token_accuracy": 0.8413710594177246, "num_tokens": 579001629.0, "step": 15180 }, { "epoch": 1.931179239282534, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.164793014526367, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8924928903579712, "num_tokens": 579036762.0, "step": 15181 }, { "epoch": 1.9313064495611245, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.184356689453125, "learning_rate": 1e-06, "loss": 0.4724, "mean_token_accuracy": 0.8462961912155151, "num_tokens": 579078455.0, "step": 15182 }, { "epoch": 1.931433659839715, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.213850021362305, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8717799186706543, "num_tokens": 579116942.0, "step": 15183 }, { "epoch": 1.9315608701183056, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.134775161743164, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8822439312934875, "num_tokens": 579155180.0, "step": 15184 }, { "epoch": 1.931688080396896, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.212193489074707, "learning_rate": 1e-06, "loss": 0.4561, "mean_token_accuracy": 0.8619696497917175, "num_tokens": 579194336.0, "step": 15185 }, { "epoch": 1.9318152906754866, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.153074264526367, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8659660816192627, "num_tokens": 579226609.0, "step": 15186 }, { "epoch": 1.9319425009540772, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.205586433410645, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8735870122909546, "num_tokens": 579259288.0, "step": 15187 }, { "epoch": 1.9320697112326677, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.14901065826416, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8776597380638123, "num_tokens": 579294797.0, "step": 15188 }, { "epoch": 1.9321969215112582, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.1422700881958, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8797625303268433, "num_tokens": 579334352.0, "step": 15189 }, { "epoch": 1.9323241317898487, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.164787292480469, "learning_rate": 1e-06, "loss": 0.472, "mean_token_accuracy": 0.8580631613731384, "num_tokens": 579379618.0, "step": 15190 }, { "epoch": 1.9324513420684393, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.2134428024292, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8687753677368164, "num_tokens": 579420230.0, "step": 15191 }, { "epoch": 1.9325785523470298, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.141338348388672, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.870418131351471, "num_tokens": 579458236.0, "step": 15192 }, { "epoch": 1.93270576262562, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.207901954650879, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8692981004714966, "num_tokens": 579495561.0, "step": 15193 }, { "epoch": 1.9328329729042106, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.184651374816895, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8683570623397827, "num_tokens": 579530225.0, "step": 15194 }, { "epoch": 1.9329601831828012, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.1810884475708, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8611900806427002, "num_tokens": 579570143.0, "step": 15195 }, { "epoch": 1.9330873934613917, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.169414520263672, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8818758130073547, "num_tokens": 579607105.0, "step": 15196 }, { "epoch": 1.9332146037399822, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.156635284423828, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8836716413497925, "num_tokens": 579640018.0, "step": 15197 }, { "epoch": 1.9333418140185727, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.204647064208984, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8638115525245667, "num_tokens": 579681250.0, "step": 15198 }, { "epoch": 1.933469024297163, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.204353332519531, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8581070899963379, "num_tokens": 579718535.0, "step": 15199 }, { "epoch": 1.9335962345757536, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.232253074645996, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8588327169418335, "num_tokens": 579756959.0, "step": 15200 }, { "epoch": 1.933723444854344, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.170515060424805, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8596597909927368, "num_tokens": 579798266.0, "step": 15201 }, { "epoch": 1.9338506551329346, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.173166275024414, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8887881636619568, "num_tokens": 579831869.0, "step": 15202 }, { "epoch": 1.9339778654115252, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.174304962158203, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8711119890213013, "num_tokens": 579874168.0, "step": 15203 }, { "epoch": 1.9341050756901157, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.154218673706055, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8750600814819336, "num_tokens": 579916313.0, "step": 15204 }, { "epoch": 1.9342322859687062, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.159074783325195, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8800733089447021, "num_tokens": 579959022.0, "step": 15205 }, { "epoch": 1.9343594962472968, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.189699172973633, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8548020124435425, "num_tokens": 580003675.0, "step": 15206 }, { "epoch": 1.9344867065258873, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.12015438079834, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8782743215560913, "num_tokens": 580044384.0, "step": 15207 }, { "epoch": 1.9346139168044778, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.237882614135742, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8585096597671509, "num_tokens": 580078955.0, "step": 15208 }, { "epoch": 1.9347411270830683, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.183176040649414, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.866073489189148, "num_tokens": 580116793.0, "step": 15209 }, { "epoch": 1.9348683373616589, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.190224647521973, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8675780296325684, "num_tokens": 580153622.0, "step": 15210 }, { "epoch": 1.9349955476402494, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.180048942565918, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8665295839309692, "num_tokens": 580195287.0, "step": 15211 }, { "epoch": 1.93512275791884, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.220295906066895, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8576446771621704, "num_tokens": 580229601.0, "step": 15212 }, { "epoch": 1.9352499681974304, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.189753532409668, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8715047836303711, "num_tokens": 580268079.0, "step": 15213 }, { "epoch": 1.935377178476021, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.19617748260498, "learning_rate": 1e-06, "loss": 0.481, "mean_token_accuracy": 0.8512004613876343, "num_tokens": 580305379.0, "step": 15214 }, { "epoch": 1.9355043887546115, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.220474243164062, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.868914783000946, "num_tokens": 580338438.0, "step": 15215 }, { "epoch": 1.935631599033202, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.199441909790039, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8697822093963623, "num_tokens": 580370017.0, "step": 15216 }, { "epoch": 1.9357588093117923, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.165045738220215, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.879535436630249, "num_tokens": 580409572.0, "step": 15217 }, { "epoch": 1.9358860195903829, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.197205543518066, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8747001886367798, "num_tokens": 580448481.0, "step": 15218 }, { "epoch": 1.9360132298689734, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.213010787963867, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8810757398605347, "num_tokens": 580481919.0, "step": 15219 }, { "epoch": 1.936140440147564, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.175202369689941, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8637984395027161, "num_tokens": 580520758.0, "step": 15220 }, { "epoch": 1.9362676504261545, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.223721504211426, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8595006465911865, "num_tokens": 580559839.0, "step": 15221 }, { "epoch": 1.936394860704745, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.20703125, "learning_rate": 1e-06, "loss": 0.3364, "mean_token_accuracy": 0.8958346247673035, "num_tokens": 580591767.0, "step": 15222 }, { "epoch": 1.9365220709833355, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.198835372924805, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.878212571144104, "num_tokens": 580633180.0, "step": 15223 }, { "epoch": 1.9366492812619258, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.213176727294922, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8718370199203491, "num_tokens": 580677045.0, "step": 15224 }, { "epoch": 1.9367764915405163, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.214255332946777, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8835869431495667, "num_tokens": 580717423.0, "step": 15225 }, { "epoch": 1.9369037018191069, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.16356372833252, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8739326000213623, "num_tokens": 580756677.0, "step": 15226 }, { "epoch": 1.9370309120976974, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.173551559448242, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.874557614326477, "num_tokens": 580796993.0, "step": 15227 }, { "epoch": 1.937158122376288, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.19231128692627, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8733237981796265, "num_tokens": 580834616.0, "step": 15228 }, { "epoch": 1.9372853326548785, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.239399909973145, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8724037408828735, "num_tokens": 580873718.0, "step": 15229 }, { "epoch": 1.937412542933469, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.189332008361816, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8654127717018127, "num_tokens": 580911678.0, "step": 15230 }, { "epoch": 1.9375397532120595, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.215518951416016, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8754138946533203, "num_tokens": 580953282.0, "step": 15231 }, { "epoch": 1.93766696349065, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.138151168823242, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.875055193901062, "num_tokens": 580997954.0, "step": 15232 }, { "epoch": 1.9377941737692406, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.2206449508667, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8705388307571411, "num_tokens": 581035623.0, "step": 15233 }, { "epoch": 1.937921384047831, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.185442924499512, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.878617525100708, "num_tokens": 581068727.0, "step": 15234 }, { "epoch": 1.9380485943264216, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.246255874633789, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8668603897094727, "num_tokens": 581111180.0, "step": 15235 }, { "epoch": 1.9381758046050122, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.171926498413086, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8496726751327515, "num_tokens": 581152373.0, "step": 15236 }, { "epoch": 1.9383030148836027, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.19698429107666, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8679242134094238, "num_tokens": 581195854.0, "step": 15237 }, { "epoch": 1.9384302251621932, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.205206871032715, "learning_rate": 1e-06, "loss": 0.5015, "mean_token_accuracy": 0.8431259989738464, "num_tokens": 581236414.0, "step": 15238 }, { "epoch": 1.9385574354407837, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.189210891723633, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.853331446647644, "num_tokens": 581275349.0, "step": 15239 }, { "epoch": 1.9386846457193743, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.170063972473145, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8769302368164062, "num_tokens": 581313231.0, "step": 15240 }, { "epoch": 1.9388118559979648, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.244626998901367, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.871784508228302, "num_tokens": 581355295.0, "step": 15241 }, { "epoch": 1.938939066276555, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.195627212524414, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8727788925170898, "num_tokens": 581392633.0, "step": 15242 }, { "epoch": 1.9390662765551456, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.200616836547852, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8838728070259094, "num_tokens": 581425552.0, "step": 15243 }, { "epoch": 1.9391934868337362, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.180445671081543, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8588311076164246, "num_tokens": 581467707.0, "step": 15244 }, { "epoch": 1.9393206971123267, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.171460151672363, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8758385181427002, "num_tokens": 581504888.0, "step": 15245 }, { "epoch": 1.9394479073909172, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.200478553771973, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8694208860397339, "num_tokens": 581537800.0, "step": 15246 }, { "epoch": 1.9395751176695077, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.19355297088623, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.8911823630332947, "num_tokens": 581570594.0, "step": 15247 }, { "epoch": 1.939702327948098, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.227228164672852, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8635357618331909, "num_tokens": 581610764.0, "step": 15248 }, { "epoch": 1.9398295382266886, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.21328353881836, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8699220418930054, "num_tokens": 581652341.0, "step": 15249 }, { "epoch": 1.939956748505279, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.195480346679688, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8611868023872375, "num_tokens": 581689978.0, "step": 15250 }, { "epoch": 1.9400839587838696, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.196614265441895, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8780220746994019, "num_tokens": 581728811.0, "step": 15251 }, { "epoch": 1.9402111690624602, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.17081069946289, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8779963850975037, "num_tokens": 581768016.0, "step": 15252 }, { "epoch": 1.9403383793410507, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.268708229064941, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8723854422569275, "num_tokens": 581805405.0, "step": 15253 }, { "epoch": 1.9404655896196412, "ewc_loss": 0.033203125, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 13.189690589904785, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8504317998886108, "num_tokens": 581845302.0, "step": 15254 }, { "epoch": 1.9405927998982317, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.249507904052734, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8793710470199585, "num_tokens": 581875391.0, "step": 15255 }, { "epoch": 1.9407200101768223, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.216273307800293, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.868913471698761, "num_tokens": 581916580.0, "step": 15256 }, { "epoch": 1.9408472204554128, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.213142395019531, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8705437183380127, "num_tokens": 581958539.0, "step": 15257 }, { "epoch": 1.9409744307340033, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.156905174255371, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8682640194892883, "num_tokens": 581996981.0, "step": 15258 }, { "epoch": 1.9411016410125939, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.183859825134277, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8791872262954712, "num_tokens": 582036068.0, "step": 15259 }, { "epoch": 1.9412288512911844, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.208885192871094, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.872421383857727, "num_tokens": 582073322.0, "step": 15260 }, { "epoch": 1.941356061569775, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.189868927001953, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8718439340591431, "num_tokens": 582110736.0, "step": 15261 }, { "epoch": 1.9414832718483654, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.1295747756958, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8533753752708435, "num_tokens": 582153431.0, "step": 15262 }, { "epoch": 1.941610482126956, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.183059692382812, "learning_rate": 1e-06, "loss": 0.4682, "mean_token_accuracy": 0.8510953783988953, "num_tokens": 582191793.0, "step": 15263 }, { "epoch": 1.9417376924055465, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.172083854675293, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8608963489532471, "num_tokens": 582227923.0, "step": 15264 }, { "epoch": 1.941864902684137, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.201637268066406, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8556361794471741, "num_tokens": 582268029.0, "step": 15265 }, { "epoch": 1.9419921129627273, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.19847297668457, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8643529415130615, "num_tokens": 582302173.0, "step": 15266 }, { "epoch": 1.9421193232413179, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.19213581085205, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8840174674987793, "num_tokens": 582341642.0, "step": 15267 }, { "epoch": 1.9422465335199084, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.247408866882324, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8718072175979614, "num_tokens": 582378960.0, "step": 15268 }, { "epoch": 1.942373743798499, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.182086944580078, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.876187264919281, "num_tokens": 582422149.0, "step": 15269 }, { "epoch": 1.9425009540770894, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.231343269348145, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.854522168636322, "num_tokens": 582462657.0, "step": 15270 }, { "epoch": 1.94262816435568, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.16831111907959, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8828316330909729, "num_tokens": 582499264.0, "step": 15271 }, { "epoch": 1.9427553746342705, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.18779182434082, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8726858496665955, "num_tokens": 582534744.0, "step": 15272 }, { "epoch": 1.9428825849128608, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.206696510314941, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8656972646713257, "num_tokens": 582577010.0, "step": 15273 }, { "epoch": 1.9430097951914513, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.136515617370605, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8649842739105225, "num_tokens": 582617103.0, "step": 15274 }, { "epoch": 1.9431370054700419, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.21960735321045, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8637696504592896, "num_tokens": 582661754.0, "step": 15275 }, { "epoch": 1.9432642157486324, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.23878002166748, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8711092472076416, "num_tokens": 582701701.0, "step": 15276 }, { "epoch": 1.943391426027223, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.204766273498535, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8616641759872437, "num_tokens": 582741977.0, "step": 15277 }, { "epoch": 1.9435186363058135, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.170788764953613, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8789904713630676, "num_tokens": 582783420.0, "step": 15278 }, { "epoch": 1.943645846584404, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.20377254486084, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8749098777770996, "num_tokens": 582819464.0, "step": 15279 }, { "epoch": 1.9437730568629945, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.167012214660645, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8758007884025574, "num_tokens": 582860757.0, "step": 15280 }, { "epoch": 1.943900267141585, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.202903747558594, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8645907044410706, "num_tokens": 582910300.0, "step": 15281 }, { "epoch": 1.9440274774201756, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.114350318908691, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8624448180198669, "num_tokens": 582956333.0, "step": 15282 }, { "epoch": 1.944154687698766, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.219088554382324, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8780676126480103, "num_tokens": 582994068.0, "step": 15283 }, { "epoch": 1.9442818979773566, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.158794403076172, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8684046864509583, "num_tokens": 583027829.0, "step": 15284 }, { "epoch": 1.9444091082559471, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.14478588104248, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8677802681922913, "num_tokens": 583068881.0, "step": 15285 }, { "epoch": 1.9445363185345377, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.238360404968262, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8735499382019043, "num_tokens": 583100097.0, "step": 15286 }, { "epoch": 1.9446635288131282, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.236075401306152, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8692891597747803, "num_tokens": 583135340.0, "step": 15287 }, { "epoch": 1.9447907390917187, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.155184745788574, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8683031797409058, "num_tokens": 583172032.0, "step": 15288 }, { "epoch": 1.9449179493703093, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.223451614379883, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8738102912902832, "num_tokens": 583205244.0, "step": 15289 }, { "epoch": 1.9450451596488998, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.21016788482666, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8665310144424438, "num_tokens": 583247537.0, "step": 15290 }, { "epoch": 1.94517236992749, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.218097686767578, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8631378412246704, "num_tokens": 583285349.0, "step": 15291 }, { "epoch": 1.9452995802060806, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.180875778198242, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8711831569671631, "num_tokens": 583323810.0, "step": 15292 }, { "epoch": 1.9454267904846712, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.193896293640137, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8828284740447998, "num_tokens": 583354244.0, "step": 15293 }, { "epoch": 1.9455540007632617, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.1554594039917, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8835147619247437, "num_tokens": 583391764.0, "step": 15294 }, { "epoch": 1.9456812110418522, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.241805076599121, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8579562306404114, "num_tokens": 583426305.0, "step": 15295 }, { "epoch": 1.9458084213204427, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.164800643920898, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8659745454788208, "num_tokens": 583465252.0, "step": 15296 }, { "epoch": 1.945935631599033, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.152875900268555, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8612774610519409, "num_tokens": 583510001.0, "step": 15297 }, { "epoch": 1.9460628418776236, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.14963436126709, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8719016313552856, "num_tokens": 583547380.0, "step": 15298 }, { "epoch": 1.946190052156214, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.18498706817627, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8709903955459595, "num_tokens": 583588852.0, "step": 15299 }, { "epoch": 1.9463172624348046, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.163484573364258, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8774037957191467, "num_tokens": 583634006.0, "step": 15300 }, { "epoch": 1.9464444727133952, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.169785499572754, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8563432693481445, "num_tokens": 583679430.0, "step": 15301 }, { "epoch": 1.9465716829919857, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.224196434020996, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.876341700553894, "num_tokens": 583708856.0, "step": 15302 }, { "epoch": 1.9466988932705762, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.161420822143555, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.868508517742157, "num_tokens": 583750737.0, "step": 15303 }, { "epoch": 1.9468261035491667, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.234332084655762, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8746232986450195, "num_tokens": 583786342.0, "step": 15304 }, { "epoch": 1.9469533138277573, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.176854133605957, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8755184412002563, "num_tokens": 583821681.0, "step": 15305 }, { "epoch": 1.9470805241063478, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.22705078125, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8520981073379517, "num_tokens": 583855640.0, "step": 15306 }, { "epoch": 1.9472077343849383, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.212817192077637, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8799906373023987, "num_tokens": 583888047.0, "step": 15307 }, { "epoch": 1.9473349446635289, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.209190368652344, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8685212135314941, "num_tokens": 583923220.0, "step": 15308 }, { "epoch": 1.9474621549421194, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.175801277160645, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8643144369125366, "num_tokens": 583965718.0, "step": 15309 }, { "epoch": 1.94758936522071, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.149848937988281, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8747473955154419, "num_tokens": 584014027.0, "step": 15310 }, { "epoch": 1.9477165754993004, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.19892406463623, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8737590312957764, "num_tokens": 584045019.0, "step": 15311 }, { "epoch": 1.947843785777891, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.191462516784668, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8524786829948425, "num_tokens": 584086839.0, "step": 15312 }, { "epoch": 1.9479709960564815, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.194029808044434, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8842592835426331, "num_tokens": 584123785.0, "step": 15313 }, { "epoch": 1.948098206335072, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.168130874633789, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8615834712982178, "num_tokens": 584163768.0, "step": 15314 }, { "epoch": 1.9482254166136623, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.239190101623535, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8664608001708984, "num_tokens": 584201668.0, "step": 15315 }, { "epoch": 1.9483526268922529, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.23745059967041, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8785291314125061, "num_tokens": 584236626.0, "step": 15316 }, { "epoch": 1.9484798371708434, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.178845405578613, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8543993234634399, "num_tokens": 584279811.0, "step": 15317 }, { "epoch": 1.948607047449434, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.194169044494629, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8696401715278625, "num_tokens": 584317314.0, "step": 15318 }, { "epoch": 1.9487342577280244, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.173635482788086, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8750330805778503, "num_tokens": 584355768.0, "step": 15319 }, { "epoch": 1.948861468006615, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.187593460083008, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8675533533096313, "num_tokens": 584386205.0, "step": 15320 }, { "epoch": 1.9489886782852053, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.246106147766113, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8818302154541016, "num_tokens": 584419721.0, "step": 15321 }, { "epoch": 1.9491158885637958, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.164937973022461, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8678500056266785, "num_tokens": 584451029.0, "step": 15322 }, { "epoch": 1.9492430988423863, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.254329681396484, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8839617371559143, "num_tokens": 584483991.0, "step": 15323 }, { "epoch": 1.9493703091209769, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.212190628051758, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8675217032432556, "num_tokens": 584517652.0, "step": 15324 }, { "epoch": 1.9494975193995674, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.151779174804688, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.877327561378479, "num_tokens": 584561058.0, "step": 15325 }, { "epoch": 1.949624729678158, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.201041221618652, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8740572929382324, "num_tokens": 584595097.0, "step": 15326 }, { "epoch": 1.9497519399567484, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.154058456420898, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8768243789672852, "num_tokens": 584639073.0, "step": 15327 }, { "epoch": 1.949879150235339, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.260331153869629, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8604648113250732, "num_tokens": 584680517.0, "step": 15328 }, { "epoch": 1.9500063605139295, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.196063995361328, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8683809041976929, "num_tokens": 584717796.0, "step": 15329 }, { "epoch": 1.95013357079252, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.210800170898438, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8828921914100647, "num_tokens": 584755595.0, "step": 15330 }, { "epoch": 1.9502607810711106, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.19234848022461, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8596872091293335, "num_tokens": 584798681.0, "step": 15331 }, { "epoch": 1.950387991349701, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.225310325622559, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8726859092712402, "num_tokens": 584835424.0, "step": 15332 }, { "epoch": 1.9505152016282916, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.193416595458984, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8836098313331604, "num_tokens": 584872063.0, "step": 15333 }, { "epoch": 1.9506424119068821, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.257624626159668, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8863568305969238, "num_tokens": 584905850.0, "step": 15334 }, { "epoch": 1.9507696221854727, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.173646926879883, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8791167140007019, "num_tokens": 584946480.0, "step": 15335 }, { "epoch": 1.9508968324640632, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.289990425109863, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8768965005874634, "num_tokens": 584977382.0, "step": 15336 }, { "epoch": 1.9510240427426537, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.196094512939453, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8723333477973938, "num_tokens": 585010885.0, "step": 15337 }, { "epoch": 1.9511512530212443, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.190261840820312, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8681639432907104, "num_tokens": 585048208.0, "step": 15338 }, { "epoch": 1.9512784632998348, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.237967491149902, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8648908138275146, "num_tokens": 585087477.0, "step": 15339 }, { "epoch": 1.951405673578425, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.198813438415527, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8695831298828125, "num_tokens": 585120860.0, "step": 15340 }, { "epoch": 1.9515328838570156, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.207674026489258, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8712830543518066, "num_tokens": 585161887.0, "step": 15341 }, { "epoch": 1.9516600941356061, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.197916030883789, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8790665864944458, "num_tokens": 585200496.0, "step": 15342 }, { "epoch": 1.9517873044141967, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.222610473632812, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8708927631378174, "num_tokens": 585243976.0, "step": 15343 }, { "epoch": 1.9519145146927872, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.20712947845459, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8652105331420898, "num_tokens": 585282366.0, "step": 15344 }, { "epoch": 1.9520417249713777, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.213288307189941, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8689053654670715, "num_tokens": 585325224.0, "step": 15345 }, { "epoch": 1.952168935249968, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.21927547454834, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.878936231136322, "num_tokens": 585361954.0, "step": 15346 }, { "epoch": 1.9522961455285586, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.238312721252441, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8671846389770508, "num_tokens": 585406033.0, "step": 15347 }, { "epoch": 1.952423355807149, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.198527336120605, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8684214353561401, "num_tokens": 585440457.0, "step": 15348 }, { "epoch": 1.9525505660857396, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.245996475219727, "learning_rate": 1e-06, "loss": 0.3474, "mean_token_accuracy": 0.8923150300979614, "num_tokens": 585473451.0, "step": 15349 }, { "epoch": 1.9526777763643302, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.220019340515137, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8746954202651978, "num_tokens": 585509652.0, "step": 15350 }, { "epoch": 1.9528049866429207, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.18410873413086, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.875115692615509, "num_tokens": 585545591.0, "step": 15351 }, { "epoch": 1.9529321969215112, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.223965644836426, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8586471080780029, "num_tokens": 585583757.0, "step": 15352 }, { "epoch": 1.9530594072001017, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.195517539978027, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8588023781776428, "num_tokens": 585623897.0, "step": 15353 }, { "epoch": 1.9531866174786923, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.245813369750977, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8738335371017456, "num_tokens": 585661043.0, "step": 15354 }, { "epoch": 1.9533138277572828, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.201116561889648, "learning_rate": 1e-06, "loss": 0.4864, "mean_token_accuracy": 0.8435117602348328, "num_tokens": 585699201.0, "step": 15355 }, { "epoch": 1.9534410380358733, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.178935050964355, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8628107309341431, "num_tokens": 585735325.0, "step": 15356 }, { "epoch": 1.9535682483144639, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.190302848815918, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8781728148460388, "num_tokens": 585769909.0, "step": 15357 }, { "epoch": 1.9536954585930544, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.191304206848145, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8584294319152832, "num_tokens": 585804590.0, "step": 15358 }, { "epoch": 1.953822668871645, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.15747356414795, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8660174608230591, "num_tokens": 585843349.0, "step": 15359 }, { "epoch": 1.9539498791502354, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.176735877990723, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8816545009613037, "num_tokens": 585881846.0, "step": 15360 }, { "epoch": 1.954077089428826, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.268576622009277, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8681368827819824, "num_tokens": 585917848.0, "step": 15361 }, { "epoch": 1.9542042997074165, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.178738594055176, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8765784502029419, "num_tokens": 585961249.0, "step": 15362 }, { "epoch": 1.954331509986007, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.242036819458008, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8782959580421448, "num_tokens": 586005795.0, "step": 15363 }, { "epoch": 1.9544587202645973, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.150805473327637, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8792983293533325, "num_tokens": 586041523.0, "step": 15364 }, { "epoch": 1.9545859305431879, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.254502296447754, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8621668815612793, "num_tokens": 586077800.0, "step": 15365 }, { "epoch": 1.9547131408217784, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.216573715209961, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8649769425392151, "num_tokens": 586116010.0, "step": 15366 }, { "epoch": 1.954840351100369, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.2310791015625, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8815566301345825, "num_tokens": 586155359.0, "step": 15367 }, { "epoch": 1.9549675613789594, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.184223175048828, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8762092590332031, "num_tokens": 586188190.0, "step": 15368 }, { "epoch": 1.95509477165755, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.235941886901855, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8698105812072754, "num_tokens": 586225971.0, "step": 15369 }, { "epoch": 1.9552219819361403, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.216211318969727, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8665733337402344, "num_tokens": 586263295.0, "step": 15370 }, { "epoch": 1.9553491922147308, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.226486206054688, "learning_rate": 1e-06, "loss": 0.3394, "mean_token_accuracy": 0.8918448090553284, "num_tokens": 586293547.0, "step": 15371 }, { "epoch": 1.9554764024933213, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.171485900878906, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8659271597862244, "num_tokens": 586333180.0, "step": 15372 }, { "epoch": 1.9556036127719119, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.223443031311035, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8711785674095154, "num_tokens": 586368780.0, "step": 15373 }, { "epoch": 1.9557308230505024, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.182703018188477, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8551672101020813, "num_tokens": 586404177.0, "step": 15374 }, { "epoch": 1.955858033329093, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.227768898010254, "learning_rate": 1e-06, "loss": 0.4671, "mean_token_accuracy": 0.8551775813102722, "num_tokens": 586446917.0, "step": 15375 }, { "epoch": 1.9559852436076834, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.242793083190918, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8708136081695557, "num_tokens": 586485832.0, "step": 15376 }, { "epoch": 1.956112453886274, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.237055778503418, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8808130025863647, "num_tokens": 586524089.0, "step": 15377 }, { "epoch": 1.9562396641648645, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.218363761901855, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8736973404884338, "num_tokens": 586564097.0, "step": 15378 }, { "epoch": 1.956366874443455, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.294241905212402, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.883558988571167, "num_tokens": 586600653.0, "step": 15379 }, { "epoch": 1.9564940847220456, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.219035148620605, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8762427568435669, "num_tokens": 586639527.0, "step": 15380 }, { "epoch": 1.956621295000636, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.251530647277832, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8752259016036987, "num_tokens": 586670885.0, "step": 15381 }, { "epoch": 1.9567485052792266, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.252558708190918, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8835647106170654, "num_tokens": 586709875.0, "step": 15382 }, { "epoch": 1.9568757155578171, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.175710678100586, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8646031618118286, "num_tokens": 586753164.0, "step": 15383 }, { "epoch": 1.9570029258364077, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.208415985107422, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8780381083488464, "num_tokens": 586793737.0, "step": 15384 }, { "epoch": 1.9571301361149982, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.242137908935547, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8661590814590454, "num_tokens": 586833428.0, "step": 15385 }, { "epoch": 1.9572573463935887, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.287429809570312, "learning_rate": 1e-06, "loss": 0.4875, "mean_token_accuracy": 0.8482383489608765, "num_tokens": 586871202.0, "step": 15386 }, { "epoch": 1.9573845566721793, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.209272384643555, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8817373514175415, "num_tokens": 586905537.0, "step": 15387 }, { "epoch": 1.9575117669507698, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.237861633300781, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8612703680992126, "num_tokens": 586940378.0, "step": 15388 }, { "epoch": 1.95763897722936, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.176695823669434, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8812448978424072, "num_tokens": 586978806.0, "step": 15389 }, { "epoch": 1.9577661875079506, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.277819633483887, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8652224540710449, "num_tokens": 587007367.0, "step": 15390 }, { "epoch": 1.9578933977865411, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.204482078552246, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.89268958568573, "num_tokens": 587040422.0, "step": 15391 }, { "epoch": 1.9580206080651317, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.177725791931152, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.887850284576416, "num_tokens": 587070238.0, "step": 15392 }, { "epoch": 1.9581478183437222, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.275796890258789, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8772381544113159, "num_tokens": 587105973.0, "step": 15393 }, { "epoch": 1.9582750286223127, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.267155647277832, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8836952447891235, "num_tokens": 587137975.0, "step": 15394 }, { "epoch": 1.958402238900903, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.255208015441895, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.866969883441925, "num_tokens": 587169176.0, "step": 15395 }, { "epoch": 1.9585294491794936, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.264900207519531, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8686197996139526, "num_tokens": 587205570.0, "step": 15396 }, { "epoch": 1.958656659458084, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.24085521697998, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8621415495872498, "num_tokens": 587246729.0, "step": 15397 }, { "epoch": 1.9587838697366746, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.215143203735352, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8755736351013184, "num_tokens": 587284537.0, "step": 15398 }, { "epoch": 1.9589110800152651, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.183507919311523, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8761235475540161, "num_tokens": 587324394.0, "step": 15399 }, { "epoch": 1.9590382902938557, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.260435104370117, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8788985013961792, "num_tokens": 587359935.0, "step": 15400 }, { "epoch": 1.9591655005724462, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.225717544555664, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8829374313354492, "num_tokens": 587402081.0, "step": 15401 }, { "epoch": 1.9592927108510367, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.219815254211426, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.868101954460144, "num_tokens": 587442939.0, "step": 15402 }, { "epoch": 1.9594199211296273, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.210722923278809, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8622563481330872, "num_tokens": 587479155.0, "step": 15403 }, { "epoch": 1.9595471314082178, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.211434364318848, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8596017360687256, "num_tokens": 587519285.0, "step": 15404 }, { "epoch": 1.9596743416868083, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.203448295593262, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8530338406562805, "num_tokens": 587563978.0, "step": 15405 }, { "epoch": 1.9598015519653988, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.279952049255371, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.865669846534729, "num_tokens": 587599052.0, "step": 15406 }, { "epoch": 1.9599287622439894, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.225647926330566, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8746728301048279, "num_tokens": 587639980.0, "step": 15407 }, { "epoch": 1.96005597252258, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.188578605651855, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8673484921455383, "num_tokens": 587673836.0, "step": 15408 }, { "epoch": 1.9601831828011704, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.210678100585938, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8719286918640137, "num_tokens": 587712896.0, "step": 15409 }, { "epoch": 1.960310393079761, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.276670455932617, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8639835715293884, "num_tokens": 587747352.0, "step": 15410 }, { "epoch": 1.9604376033583515, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.27114200592041, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8696587085723877, "num_tokens": 587790234.0, "step": 15411 }, { "epoch": 1.960564813636942, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.224385261535645, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8845194578170776, "num_tokens": 587830185.0, "step": 15412 }, { "epoch": 1.9606920239155323, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.272031784057617, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8733170032501221, "num_tokens": 587866265.0, "step": 15413 }, { "epoch": 1.9608192341941229, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.206197738647461, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8831408619880676, "num_tokens": 587904956.0, "step": 15414 }, { "epoch": 1.9609464444727134, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.192093849182129, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.878250241279602, "num_tokens": 587942369.0, "step": 15415 }, { "epoch": 1.961073654751304, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.212339401245117, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8756823539733887, "num_tokens": 587979340.0, "step": 15416 }, { "epoch": 1.9612008650298944, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.21435546875, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8666054010391235, "num_tokens": 588021812.0, "step": 15417 }, { "epoch": 1.961328075308485, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.194963455200195, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8672839403152466, "num_tokens": 588064978.0, "step": 15418 }, { "epoch": 1.9614552855870753, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.27790355682373, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8664018511772156, "num_tokens": 588098142.0, "step": 15419 }, { "epoch": 1.9615824958656658, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.144250869750977, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8776753544807434, "num_tokens": 588141808.0, "step": 15420 }, { "epoch": 1.9617097061442563, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.170453071594238, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8779666423797607, "num_tokens": 588178137.0, "step": 15421 }, { "epoch": 1.9618369164228469, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.248197555541992, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8650761842727661, "num_tokens": 588215646.0, "step": 15422 }, { "epoch": 1.9619641267014374, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.213530540466309, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8621994256973267, "num_tokens": 588251345.0, "step": 15423 }, { "epoch": 1.962091336980028, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.22679615020752, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8774595856666565, "num_tokens": 588291222.0, "step": 15424 }, { "epoch": 1.9622185472586184, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.23996639251709, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8625442981719971, "num_tokens": 588329480.0, "step": 15425 }, { "epoch": 1.962345757537209, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.243426322937012, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8623915910720825, "num_tokens": 588367143.0, "step": 15426 }, { "epoch": 1.9624729678157995, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.196115493774414, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8789419531822205, "num_tokens": 588408363.0, "step": 15427 }, { "epoch": 1.96260017809439, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.259220123291016, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8775644302368164, "num_tokens": 588445397.0, "step": 15428 }, { "epoch": 1.9627273883729806, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.237449645996094, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8769473433494568, "num_tokens": 588480436.0, "step": 15429 }, { "epoch": 1.962854598651571, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.21269702911377, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8693498373031616, "num_tokens": 588526918.0, "step": 15430 }, { "epoch": 1.9629818089301616, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.165510177612305, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.871645450592041, "num_tokens": 588561300.0, "step": 15431 }, { "epoch": 1.9631090192087521, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.277048110961914, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8615413904190063, "num_tokens": 588603393.0, "step": 15432 }, { "epoch": 1.9632362294873427, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.23030948638916, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8780618906021118, "num_tokens": 588643806.0, "step": 15433 }, { "epoch": 1.9633634397659332, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.24341106414795, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.870313286781311, "num_tokens": 588685090.0, "step": 15434 }, { "epoch": 1.9634906500445237, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 13.218640327453613, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8591123819351196, "num_tokens": 588720983.0, "step": 15435 }, { "epoch": 1.9636178603231143, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.225045204162598, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8637710809707642, "num_tokens": 588760143.0, "step": 15436 }, { "epoch": 1.9637450706017048, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.282833099365234, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8730119466781616, "num_tokens": 588796636.0, "step": 15437 }, { "epoch": 1.963872280880295, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.212701797485352, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8677902221679688, "num_tokens": 588831681.0, "step": 15438 }, { "epoch": 1.9639994911588856, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.236981391906738, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8811715245246887, "num_tokens": 588875408.0, "step": 15439 }, { "epoch": 1.9641267014374761, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.228497505187988, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8748948574066162, "num_tokens": 588913583.0, "step": 15440 }, { "epoch": 1.9642539117160667, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.282776832580566, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8801844716072083, "num_tokens": 588949833.0, "step": 15441 }, { "epoch": 1.9643811219946572, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.248472213745117, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8764661550521851, "num_tokens": 588985678.0, "step": 15442 }, { "epoch": 1.9645083322732477, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.296126365661621, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8705164790153503, "num_tokens": 589018276.0, "step": 15443 }, { "epoch": 1.964635542551838, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.23141860961914, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8697733283042908, "num_tokens": 589061900.0, "step": 15444 }, { "epoch": 1.9647627528304286, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.25847053527832, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8736066818237305, "num_tokens": 589102740.0, "step": 15445 }, { "epoch": 1.964889963109019, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.262916564941406, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8803799152374268, "num_tokens": 589135276.0, "step": 15446 }, { "epoch": 1.9650171733876096, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.22729778289795, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8710935711860657, "num_tokens": 589173390.0, "step": 15447 }, { "epoch": 1.9651443836662001, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.281623840332031, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8643103837966919, "num_tokens": 589214592.0, "step": 15448 }, { "epoch": 1.9652715939447907, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.307433128356934, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8711869716644287, "num_tokens": 589250682.0, "step": 15449 }, { "epoch": 1.9653988042233812, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.26504898071289, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.864795982837677, "num_tokens": 589288326.0, "step": 15450 }, { "epoch": 1.9655260145019717, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.286652565002441, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8674973845481873, "num_tokens": 589321632.0, "step": 15451 }, { "epoch": 1.9656532247805623, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.25938892364502, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8616845607757568, "num_tokens": 589361134.0, "step": 15452 }, { "epoch": 1.9657804350591528, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.264257431030273, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8731784820556641, "num_tokens": 589402338.0, "step": 15453 }, { "epoch": 1.9659076453377433, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.197857856750488, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8656480312347412, "num_tokens": 589441080.0, "step": 15454 }, { "epoch": 1.9660348556163338, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.260303497314453, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8678942322731018, "num_tokens": 589484019.0, "step": 15455 }, { "epoch": 1.9661620658949244, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.25387191772461, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8663012981414795, "num_tokens": 589525890.0, "step": 15456 }, { "epoch": 1.966289276173515, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.26818561553955, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8761729598045349, "num_tokens": 589560644.0, "step": 15457 }, { "epoch": 1.9664164864521054, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.269103050231934, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8907773494720459, "num_tokens": 589600863.0, "step": 15458 }, { "epoch": 1.966543696730696, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.236529350280762, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8704699277877808, "num_tokens": 589637948.0, "step": 15459 }, { "epoch": 1.9666709070092865, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.22983169555664, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8836205005645752, "num_tokens": 589678061.0, "step": 15460 }, { "epoch": 1.966798117287877, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.331038475036621, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8598086833953857, "num_tokens": 589718050.0, "step": 15461 }, { "epoch": 1.9669253275664673, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.247350692749023, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8739778399467468, "num_tokens": 589759206.0, "step": 15462 }, { "epoch": 1.9670525378450578, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.29249382019043, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8594685792922974, "num_tokens": 589800218.0, "step": 15463 }, { "epoch": 1.9671797481236484, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.290318489074707, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8719601631164551, "num_tokens": 589843105.0, "step": 15464 }, { "epoch": 1.967306958402239, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.249027252197266, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8626034259796143, "num_tokens": 589881560.0, "step": 15465 }, { "epoch": 1.9674341686808294, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.282663345336914, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8590517640113831, "num_tokens": 589922755.0, "step": 15466 }, { "epoch": 1.96756137895942, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.27531909942627, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.852218508720398, "num_tokens": 589964982.0, "step": 15467 }, { "epoch": 1.9676885892380103, "ewc_loss": 0.033447265625, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.262266159057617, "learning_rate": 1e-06, "loss": 0.455, "mean_token_accuracy": 0.8571144342422485, "num_tokens": 590003371.0, "step": 15468 }, { "epoch": 1.9678157995166008, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.160017967224121, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8686744570732117, "num_tokens": 590048361.0, "step": 15469 }, { "epoch": 1.9679430097951913, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.32740592956543, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8690873980522156, "num_tokens": 590083141.0, "step": 15470 }, { "epoch": 1.9680702200737819, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.20206069946289, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8646982312202454, "num_tokens": 590128161.0, "step": 15471 }, { "epoch": 1.9681974303523724, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.2816801071167, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8857852816581726, "num_tokens": 590166101.0, "step": 15472 }, { "epoch": 1.968324640630963, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.26161003112793, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8696125149726868, "num_tokens": 590207643.0, "step": 15473 }, { "epoch": 1.9684518509095534, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.258684158325195, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8630834817886353, "num_tokens": 590247567.0, "step": 15474 }, { "epoch": 1.968579061188144, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.297038078308105, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8668233156204224, "num_tokens": 590287940.0, "step": 15475 }, { "epoch": 1.9687062714667345, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.261980056762695, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8641432523727417, "num_tokens": 590322558.0, "step": 15476 }, { "epoch": 1.968833481745325, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.261817932128906, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8689950704574585, "num_tokens": 590362913.0, "step": 15477 }, { "epoch": 1.9689606920239155, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.255088806152344, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8620362877845764, "num_tokens": 590405877.0, "step": 15478 }, { "epoch": 1.969087902302506, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.304525375366211, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.8612158298492432, "num_tokens": 590445373.0, "step": 15479 }, { "epoch": 1.9692151125810966, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.285614967346191, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.8595545887947083, "num_tokens": 590483319.0, "step": 15480 }, { "epoch": 1.9693423228596871, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.284736633300781, "learning_rate": 1e-06, "loss": 0.4593, "mean_token_accuracy": 0.8564584255218506, "num_tokens": 590522433.0, "step": 15481 }, { "epoch": 1.9694695331382777, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.32730484008789, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8741545677185059, "num_tokens": 590564342.0, "step": 15482 }, { "epoch": 1.9695967434168682, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.258628845214844, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8726255893707275, "num_tokens": 590594527.0, "step": 15483 }, { "epoch": 1.9697239536954587, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.308456420898438, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.88194739818573, "num_tokens": 590630573.0, "step": 15484 }, { "epoch": 1.9698511639740492, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.188992500305176, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8714125752449036, "num_tokens": 590669333.0, "step": 15485 }, { "epoch": 1.9699783742526398, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.295380592346191, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8859951496124268, "num_tokens": 590709884.0, "step": 15486 }, { "epoch": 1.97010558453123, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.316336631774902, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8758838176727295, "num_tokens": 590746109.0, "step": 15487 }, { "epoch": 1.9702327948098206, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.23729133605957, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8547461032867432, "num_tokens": 590787881.0, "step": 15488 }, { "epoch": 1.9703600050884111, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.238558769226074, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8560351133346558, "num_tokens": 590826861.0, "step": 15489 }, { "epoch": 1.9704872153670017, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.243223190307617, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8787156343460083, "num_tokens": 590863104.0, "step": 15490 }, { "epoch": 1.9706144256455922, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.275236129760742, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8774845600128174, "num_tokens": 590900582.0, "step": 15491 }, { "epoch": 1.9707416359241827, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.26950454711914, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8768335580825806, "num_tokens": 590937735.0, "step": 15492 }, { "epoch": 1.970868846202773, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.202625274658203, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8842360973358154, "num_tokens": 590973089.0, "step": 15493 }, { "epoch": 1.9709960564813636, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.287070274353027, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8573389053344727, "num_tokens": 591012117.0, "step": 15494 }, { "epoch": 1.971123266759954, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.22125244140625, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.8515209555625916, "num_tokens": 591044978.0, "step": 15495 }, { "epoch": 1.9712504770385446, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.241575241088867, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8668025135993958, "num_tokens": 591085562.0, "step": 15496 }, { "epoch": 1.9713776873171351, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.198403358459473, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8611289858818054, "num_tokens": 591120140.0, "step": 15497 }, { "epoch": 1.9715048975957257, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.293784141540527, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8605650663375854, "num_tokens": 591153952.0, "step": 15498 }, { "epoch": 1.9716321078743162, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.186796188354492, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8864309787750244, "num_tokens": 591192983.0, "step": 15499 }, { "epoch": 1.9717593181529067, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.27050495147705, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8716269731521606, "num_tokens": 591226260.0, "step": 15500 }, { "epoch": 1.9718865284314973, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.250248908996582, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.857768177986145, "num_tokens": 591259361.0, "step": 15501 }, { "epoch": 1.9720137387100878, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.264105796813965, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8705349564552307, "num_tokens": 591299753.0, "step": 15502 }, { "epoch": 1.9721409489886783, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.246885299682617, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8716592788696289, "num_tokens": 591335008.0, "step": 15503 }, { "epoch": 1.9722681592672688, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.244723320007324, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.877468466758728, "num_tokens": 591373001.0, "step": 15504 }, { "epoch": 1.9723953695458594, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.258868217468262, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8726937770843506, "num_tokens": 591408637.0, "step": 15505 }, { "epoch": 1.97252257982445, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.269906044006348, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8671907186508179, "num_tokens": 591444908.0, "step": 15506 }, { "epoch": 1.9726497901030404, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.28592300415039, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8726301193237305, "num_tokens": 591484641.0, "step": 15507 }, { "epoch": 1.972777000381631, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.234704971313477, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8668297529220581, "num_tokens": 591521859.0, "step": 15508 }, { "epoch": 1.9729042106602215, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.218844413757324, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8695982098579407, "num_tokens": 591559124.0, "step": 15509 }, { "epoch": 1.973031420938812, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.273378372192383, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8689061999320984, "num_tokens": 591597170.0, "step": 15510 }, { "epoch": 1.9731586312174023, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.2744722366333, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8760628700256348, "num_tokens": 591634784.0, "step": 15511 }, { "epoch": 1.9732858414959928, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.219829559326172, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8742311000823975, "num_tokens": 591671904.0, "step": 15512 }, { "epoch": 1.9734130517745834, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.242088317871094, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8715825080871582, "num_tokens": 591712990.0, "step": 15513 }, { "epoch": 1.973540262053174, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.31362533569336, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8669579029083252, "num_tokens": 591748579.0, "step": 15514 }, { "epoch": 1.9736674723317644, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.260528564453125, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8638615608215332, "num_tokens": 591785865.0, "step": 15515 }, { "epoch": 1.973794682610355, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.279417037963867, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8791478872299194, "num_tokens": 591820087.0, "step": 15516 }, { "epoch": 1.9739218928889453, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.307284355163574, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8691589832305908, "num_tokens": 591859713.0, "step": 15517 }, { "epoch": 1.9740491031675358, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.21727466583252, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8771984577178955, "num_tokens": 591893142.0, "step": 15518 }, { "epoch": 1.9741763134461263, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.289071083068848, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8655121326446533, "num_tokens": 591931529.0, "step": 15519 }, { "epoch": 1.9743035237247168, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.263568878173828, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8753794431686401, "num_tokens": 591976248.0, "step": 15520 }, { "epoch": 1.9744307340033074, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.265774726867676, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.867916464805603, "num_tokens": 592018008.0, "step": 15521 }, { "epoch": 1.974557944281898, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.244625091552734, "learning_rate": 1e-06, "loss": 0.4583, "mean_token_accuracy": 0.8554093837738037, "num_tokens": 592062934.0, "step": 15522 }, { "epoch": 1.9746851545604884, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.25606918334961, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.868516743183136, "num_tokens": 592103129.0, "step": 15523 }, { "epoch": 1.974812364839079, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.271197319030762, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8656148910522461, "num_tokens": 592142116.0, "step": 15524 }, { "epoch": 1.9749395751176695, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.268033027648926, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8707824945449829, "num_tokens": 592182507.0, "step": 15525 }, { "epoch": 1.97506678539626, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.241046905517578, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8483099937438965, "num_tokens": 592225593.0, "step": 15526 }, { "epoch": 1.9751939956748505, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.275880813598633, "learning_rate": 1e-06, "loss": 0.3393, "mean_token_accuracy": 0.8942509293556213, "num_tokens": 592270676.0, "step": 15527 }, { "epoch": 1.975321205953441, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.276726722717285, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.888342022895813, "num_tokens": 592305537.0, "step": 15528 }, { "epoch": 1.9754484162320316, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.261346817016602, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8769466876983643, "num_tokens": 592338289.0, "step": 15529 }, { "epoch": 1.9755756265106221, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.276620864868164, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.873225212097168, "num_tokens": 592379189.0, "step": 15530 }, { "epoch": 1.9757028367892127, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.228692054748535, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.884296178817749, "num_tokens": 592421875.0, "step": 15531 }, { "epoch": 1.9758300470678032, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.325823783874512, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8512890338897705, "num_tokens": 592457438.0, "step": 15532 }, { "epoch": 1.9759572573463937, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.28378963470459, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8812775611877441, "num_tokens": 592493230.0, "step": 15533 }, { "epoch": 1.9760844676249842, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.263778686523438, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8542163968086243, "num_tokens": 592533656.0, "step": 15534 }, { "epoch": 1.9762116779035748, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.210711479187012, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8833200931549072, "num_tokens": 592571622.0, "step": 15535 }, { "epoch": 1.976338888182165, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.207100868225098, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8649111390113831, "num_tokens": 592613260.0, "step": 15536 }, { "epoch": 1.9764660984607556, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.296829223632812, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8679422736167908, "num_tokens": 592648872.0, "step": 15537 }, { "epoch": 1.9765933087393461, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.264245986938477, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.871079683303833, "num_tokens": 592682336.0, "step": 15538 }, { "epoch": 1.9767205190179367, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.27979850769043, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8847667574882507, "num_tokens": 592716412.0, "step": 15539 }, { "epoch": 1.9768477292965272, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.264317512512207, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.873641848564148, "num_tokens": 592753936.0, "step": 15540 }, { "epoch": 1.9769749395751177, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.311335563659668, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8710060119628906, "num_tokens": 592792731.0, "step": 15541 }, { "epoch": 1.977102149853708, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.243083000183105, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8739051818847656, "num_tokens": 592832035.0, "step": 15542 }, { "epoch": 1.9772293601322986, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.305359840393066, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.887836754322052, "num_tokens": 592869693.0, "step": 15543 }, { "epoch": 1.977356570410889, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.27847957611084, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8738930821418762, "num_tokens": 592909648.0, "step": 15544 }, { "epoch": 1.9774837806894796, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.291547775268555, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8739079236984253, "num_tokens": 592942539.0, "step": 15545 }, { "epoch": 1.9776109909680701, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.29967212677002, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8693996667861938, "num_tokens": 592983680.0, "step": 15546 }, { "epoch": 1.9777382012466607, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.303421974182129, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8594638109207153, "num_tokens": 593023161.0, "step": 15547 }, { "epoch": 1.9778654115252512, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.266683578491211, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8519554138183594, "num_tokens": 593066612.0, "step": 15548 }, { "epoch": 1.9779926218038417, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.307613372802734, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8584721088409424, "num_tokens": 593105224.0, "step": 15549 }, { "epoch": 1.9781198320824323, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.251707077026367, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8710570335388184, "num_tokens": 593143683.0, "step": 15550 }, { "epoch": 1.9782470423610228, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.31851577758789, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8563976883888245, "num_tokens": 593179827.0, "step": 15551 }, { "epoch": 1.9783742526396133, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.197647094726562, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8712896108627319, "num_tokens": 593220936.0, "step": 15552 }, { "epoch": 1.9785014629182038, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.250861167907715, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8659189343452454, "num_tokens": 593258503.0, "step": 15553 }, { "epoch": 1.9786286731967944, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.272176742553711, "learning_rate": 1e-06, "loss": 0.4932, "mean_token_accuracy": 0.8512627482414246, "num_tokens": 593293738.0, "step": 15554 }, { "epoch": 1.978755883475385, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.24946403503418, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8609662055969238, "num_tokens": 593338037.0, "step": 15555 }, { "epoch": 1.9788830937539754, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.227720260620117, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8545105457305908, "num_tokens": 593383509.0, "step": 15556 }, { "epoch": 1.979010304032566, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.259197235107422, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8741121292114258, "num_tokens": 593417734.0, "step": 15557 }, { "epoch": 1.9791375143111565, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.231687545776367, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.882183849811554, "num_tokens": 593454502.0, "step": 15558 }, { "epoch": 1.979264724589747, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.30471420288086, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8663507699966431, "num_tokens": 593498864.0, "step": 15559 }, { "epoch": 1.9793919348683373, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.260197639465332, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8663005232810974, "num_tokens": 593540125.0, "step": 15560 }, { "epoch": 1.9795191451469278, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.28654670715332, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8805307149887085, "num_tokens": 593566937.0, "step": 15561 }, { "epoch": 1.9796463554255184, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.259371757507324, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8657701015472412, "num_tokens": 593600109.0, "step": 15562 }, { "epoch": 1.979773565704109, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.25670051574707, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8688403367996216, "num_tokens": 593638982.0, "step": 15563 }, { "epoch": 1.9799007759826994, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.268949508666992, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8601242303848267, "num_tokens": 593677270.0, "step": 15564 }, { "epoch": 1.98002798626129, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.262077331542969, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8769112825393677, "num_tokens": 593714056.0, "step": 15565 }, { "epoch": 1.9801551965398803, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.268760681152344, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8591921329498291, "num_tokens": 593753516.0, "step": 15566 }, { "epoch": 1.9802824068184708, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.262850761413574, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8724834322929382, "num_tokens": 593789276.0, "step": 15567 }, { "epoch": 1.9804096170970613, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.25541877746582, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8743502497673035, "num_tokens": 593828112.0, "step": 15568 }, { "epoch": 1.9805368273756518, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.266331672668457, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8579798936843872, "num_tokens": 593866370.0, "step": 15569 }, { "epoch": 1.9806640376542424, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.268214225769043, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8745510578155518, "num_tokens": 593907863.0, "step": 15570 }, { "epoch": 1.980791247932833, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.260377883911133, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8810456991195679, "num_tokens": 593945254.0, "step": 15571 }, { "epoch": 1.9809184582114234, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.273696899414062, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8736995458602905, "num_tokens": 593987105.0, "step": 15572 }, { "epoch": 1.981045668490014, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.25502872467041, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.881201982498169, "num_tokens": 594024091.0, "step": 15573 }, { "epoch": 1.9811728787686045, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.298284530639648, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8811957836151123, "num_tokens": 594056018.0, "step": 15574 }, { "epoch": 1.981300089047195, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.227232933044434, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8726712465286255, "num_tokens": 594102297.0, "step": 15575 }, { "epoch": 1.9814272993257855, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.230644226074219, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8713297843933105, "num_tokens": 594140278.0, "step": 15576 }, { "epoch": 1.981554509604376, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.282312393188477, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8747771978378296, "num_tokens": 594175025.0, "step": 15577 }, { "epoch": 1.9816817198829666, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.216629028320312, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8691575527191162, "num_tokens": 594217906.0, "step": 15578 }, { "epoch": 1.9818089301615571, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.219348907470703, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.853448748588562, "num_tokens": 594264331.0, "step": 15579 }, { "epoch": 1.9819361404401477, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.214725494384766, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8854104280471802, "num_tokens": 594300220.0, "step": 15580 }, { "epoch": 1.9820633507187382, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.286972999572754, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8760636448860168, "num_tokens": 594337849.0, "step": 15581 }, { "epoch": 1.9821905609973287, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.253527641296387, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8645442724227905, "num_tokens": 594379295.0, "step": 15582 }, { "epoch": 1.9823177712759192, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.271456718444824, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8669103384017944, "num_tokens": 594408350.0, "step": 15583 }, { "epoch": 1.9824449815545098, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.264198303222656, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.859725296497345, "num_tokens": 594455110.0, "step": 15584 }, { "epoch": 1.9825721918331, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.284646034240723, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8743979930877686, "num_tokens": 594499231.0, "step": 15585 }, { "epoch": 1.9826994021116906, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.299298286437988, "learning_rate": 1e-06, "loss": 0.4753, "mean_token_accuracy": 0.8510897159576416, "num_tokens": 594541478.0, "step": 15586 }, { "epoch": 1.9828266123902811, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.284666061401367, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8728626370429993, "num_tokens": 594579075.0, "step": 15587 }, { "epoch": 1.9829538226688717, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.290931701660156, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8862791657447815, "num_tokens": 594616189.0, "step": 15588 }, { "epoch": 1.9830810329474622, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.260213851928711, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8772203922271729, "num_tokens": 594656633.0, "step": 15589 }, { "epoch": 1.9832082432260527, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.255008697509766, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8519688844680786, "num_tokens": 594696833.0, "step": 15590 }, { "epoch": 1.983335453504643, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.271943092346191, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8573932647705078, "num_tokens": 594735138.0, "step": 15591 }, { "epoch": 1.9834626637832335, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.271169662475586, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8644157648086548, "num_tokens": 594769587.0, "step": 15592 }, { "epoch": 1.983589874061824, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.291637420654297, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8759011626243591, "num_tokens": 594803635.0, "step": 15593 }, { "epoch": 1.9837170843404146, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.315957069396973, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8865352869033813, "num_tokens": 594839872.0, "step": 15594 }, { "epoch": 1.9838442946190051, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.319908142089844, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8584067225456238, "num_tokens": 594879208.0, "step": 15595 }, { "epoch": 1.9839715048975957, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.29818344116211, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.876713752746582, "num_tokens": 594910377.0, "step": 15596 }, { "epoch": 1.9840987151761862, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.308382034301758, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8727349638938904, "num_tokens": 594956574.0, "step": 15597 }, { "epoch": 1.9842259254547767, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.245418548583984, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8547523617744446, "num_tokens": 594998036.0, "step": 15598 }, { "epoch": 1.9843531357333672, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.361139297485352, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.865890622138977, "num_tokens": 595031404.0, "step": 15599 }, { "epoch": 1.9844803460119578, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.253148078918457, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8788619041442871, "num_tokens": 595069800.0, "step": 15600 }, { "epoch": 1.9846075562905483, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.339502334594727, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8754358291625977, "num_tokens": 595110987.0, "step": 15601 }, { "epoch": 1.9847347665691388, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.257902145385742, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8738343715667725, "num_tokens": 595149192.0, "step": 15602 }, { "epoch": 1.9848619768477294, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.320240020751953, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8688395619392395, "num_tokens": 595182993.0, "step": 15603 }, { "epoch": 1.9849891871263199, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.269648551940918, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8668497800827026, "num_tokens": 595216177.0, "step": 15604 }, { "epoch": 1.9851163974049104, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.313030242919922, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8716169595718384, "num_tokens": 595253379.0, "step": 15605 }, { "epoch": 1.985243607683501, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.224705696105957, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8749197721481323, "num_tokens": 595292878.0, "step": 15606 }, { "epoch": 1.9853708179620915, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.26768684387207, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8719395399093628, "num_tokens": 595329082.0, "step": 15607 }, { "epoch": 1.985498028240682, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.303418159484863, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8632543683052063, "num_tokens": 595367246.0, "step": 15608 }, { "epoch": 1.9856252385192723, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.2405366897583, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8775111436843872, "num_tokens": 595404450.0, "step": 15609 }, { "epoch": 1.9857524487978628, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.294730186462402, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8833493590354919, "num_tokens": 595439718.0, "step": 15610 }, { "epoch": 1.9858796590764534, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.272253036499023, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8731681108474731, "num_tokens": 595473617.0, "step": 15611 }, { "epoch": 1.986006869355044, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.28549861907959, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8682640790939331, "num_tokens": 595516432.0, "step": 15612 }, { "epoch": 1.9861340796336344, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.281538963317871, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8710346221923828, "num_tokens": 595556278.0, "step": 15613 }, { "epoch": 1.986261289912225, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.330382347106934, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.876093864440918, "num_tokens": 595598713.0, "step": 15614 }, { "epoch": 1.9863885001908153, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.334278106689453, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8737680912017822, "num_tokens": 595630597.0, "step": 15615 }, { "epoch": 1.9865157104694058, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.256077766418457, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8881881833076477, "num_tokens": 595664163.0, "step": 15616 }, { "epoch": 1.9866429207479963, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.284320831298828, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8797577619552612, "num_tokens": 595698809.0, "step": 15617 }, { "epoch": 1.9867701310265868, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.31666088104248, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8652092814445496, "num_tokens": 595741166.0, "step": 15618 }, { "epoch": 1.9868973413051774, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.270726203918457, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8664067983627319, "num_tokens": 595787079.0, "step": 15619 }, { "epoch": 1.987024551583768, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.29000186920166, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.8455742001533508, "num_tokens": 595829085.0, "step": 15620 }, { "epoch": 1.9871517618623584, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.24617862701416, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8694968223571777, "num_tokens": 595870646.0, "step": 15621 }, { "epoch": 1.987278972140949, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.27245044708252, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8604588508605957, "num_tokens": 595918473.0, "step": 15622 }, { "epoch": 1.9874061824195395, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.334657669067383, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8609634637832642, "num_tokens": 595954097.0, "step": 15623 }, { "epoch": 1.98753339269813, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.26231861114502, "learning_rate": 1e-06, "loss": 0.4635, "mean_token_accuracy": 0.8520539999008179, "num_tokens": 595990191.0, "step": 15624 }, { "epoch": 1.9876606029767205, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.258298873901367, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8635609149932861, "num_tokens": 596027109.0, "step": 15625 }, { "epoch": 1.987787813255311, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.330950736999512, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8783398270606995, "num_tokens": 596067930.0, "step": 15626 }, { "epoch": 1.9879150235339016, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.282901763916016, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8807239532470703, "num_tokens": 596105809.0, "step": 15627 }, { "epoch": 1.9880422338124921, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.354172706604004, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8743828535079956, "num_tokens": 596144220.0, "step": 15628 }, { "epoch": 1.9881694440910826, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.34862232208252, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.848236083984375, "num_tokens": 596183825.0, "step": 15629 }, { "epoch": 1.9882966543696732, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.307454109191895, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8786695003509521, "num_tokens": 596232333.0, "step": 15630 }, { "epoch": 1.9884238646482637, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.284982681274414, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8783629536628723, "num_tokens": 596272138.0, "step": 15631 }, { "epoch": 1.9885510749268542, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.342016220092773, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8796266317367554, "num_tokens": 596310254.0, "step": 15632 }, { "epoch": 1.9886782852054448, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.172518730163574, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8678234815597534, "num_tokens": 596350396.0, "step": 15633 }, { "epoch": 1.988805495484035, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.305184364318848, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8749431371688843, "num_tokens": 596388809.0, "step": 15634 }, { "epoch": 1.9889327057626256, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.304430961608887, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8787075281143188, "num_tokens": 596427366.0, "step": 15635 }, { "epoch": 1.9890599160412161, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.269408226013184, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8610537052154541, "num_tokens": 596469999.0, "step": 15636 }, { "epoch": 1.9891871263198067, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.281059265136719, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8553764224052429, "num_tokens": 596506773.0, "step": 15637 }, { "epoch": 1.9893143365983972, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.260342597961426, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8718913197517395, "num_tokens": 596547440.0, "step": 15638 }, { "epoch": 1.9894415468769877, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.273958206176758, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8770896196365356, "num_tokens": 596580212.0, "step": 15639 }, { "epoch": 1.989568757155578, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.302855491638184, "learning_rate": 1e-06, "loss": 0.3466, "mean_token_accuracy": 0.8920964002609253, "num_tokens": 596611790.0, "step": 15640 }, { "epoch": 1.9896959674341685, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.291707992553711, "learning_rate": 1e-06, "loss": 0.3464, "mean_token_accuracy": 0.8861045837402344, "num_tokens": 596646984.0, "step": 15641 }, { "epoch": 1.989823177712759, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.233724594116211, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8666689395904541, "num_tokens": 596688362.0, "step": 15642 }, { "epoch": 1.9899503879913496, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.290817260742188, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8737879395484924, "num_tokens": 596725298.0, "step": 15643 }, { "epoch": 1.9900775982699401, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.269758224487305, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.877479612827301, "num_tokens": 596757400.0, "step": 15644 }, { "epoch": 1.9902048085485307, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.25549030303955, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.88239586353302, "num_tokens": 596790533.0, "step": 15645 }, { "epoch": 1.9903320188271212, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.276987075805664, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8648996949195862, "num_tokens": 596834467.0, "step": 15646 }, { "epoch": 1.9904592291057117, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.227363586425781, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8590102195739746, "num_tokens": 596873262.0, "step": 15647 }, { "epoch": 1.9905864393843022, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.295226097106934, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8711333274841309, "num_tokens": 596912833.0, "step": 15648 }, { "epoch": 1.9907136496628928, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.248720169067383, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8788046836853027, "num_tokens": 596953554.0, "step": 15649 }, { "epoch": 1.9908408599414833, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.239550590515137, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8652178645133972, "num_tokens": 596991206.0, "step": 15650 }, { "epoch": 1.9909680702200738, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.314671516418457, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8721246719360352, "num_tokens": 597027222.0, "step": 15651 }, { "epoch": 1.9910952804986644, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.205326080322266, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.879460871219635, "num_tokens": 597069040.0, "step": 15652 }, { "epoch": 1.9912224907772549, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.260244369506836, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8636157512664795, "num_tokens": 597112529.0, "step": 15653 }, { "epoch": 1.9913497010558454, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.33021068572998, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8730841875076294, "num_tokens": 597148304.0, "step": 15654 }, { "epoch": 1.991476911334436, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.222966194152832, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8791037201881409, "num_tokens": 597184244.0, "step": 15655 }, { "epoch": 1.9916041216130265, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.322042465209961, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8638893365859985, "num_tokens": 597218178.0, "step": 15656 }, { "epoch": 1.991731331891617, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.303864479064941, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8688486814498901, "num_tokens": 597259256.0, "step": 15657 }, { "epoch": 1.9918585421702073, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.264115333557129, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8815621733665466, "num_tokens": 597296584.0, "step": 15658 }, { "epoch": 1.9919857524487978, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.332014083862305, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8563947081565857, "num_tokens": 597341116.0, "step": 15659 }, { "epoch": 1.9921129627273884, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.27245044708252, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8681648373603821, "num_tokens": 597375595.0, "step": 15660 }, { "epoch": 1.9922401730059789, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.297948837280273, "learning_rate": 1e-06, "loss": 0.4525, "mean_token_accuracy": 0.8573960661888123, "num_tokens": 597419226.0, "step": 15661 }, { "epoch": 1.9923673832845694, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.335528373718262, "learning_rate": 1e-06, "loss": 0.4686, "mean_token_accuracy": 0.8518640995025635, "num_tokens": 597461242.0, "step": 15662 }, { "epoch": 1.99249459356316, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.277325630187988, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8702400922775269, "num_tokens": 597500030.0, "step": 15663 }, { "epoch": 1.9926218038417502, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.31105899810791, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8559888601303101, "num_tokens": 597540428.0, "step": 15664 }, { "epoch": 1.9927490141203408, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.296046257019043, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8807826638221741, "num_tokens": 597575070.0, "step": 15665 }, { "epoch": 1.9928762243989313, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.278181076049805, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8696739077568054, "num_tokens": 597617425.0, "step": 15666 }, { "epoch": 1.9930034346775218, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.237649917602539, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8778349757194519, "num_tokens": 597660489.0, "step": 15667 }, { "epoch": 1.9931306449561124, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.269620895385742, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8760780096054077, "num_tokens": 597703626.0, "step": 15668 }, { "epoch": 1.993257855234703, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.261168479919434, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8671092987060547, "num_tokens": 597741176.0, "step": 15669 }, { "epoch": 1.9933850655132934, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.304241180419922, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8583357334136963, "num_tokens": 597779763.0, "step": 15670 }, { "epoch": 1.993512275791884, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.28837776184082, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8736966252326965, "num_tokens": 597813300.0, "step": 15671 }, { "epoch": 1.9936394860704745, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.285143852233887, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8604966998100281, "num_tokens": 597849062.0, "step": 15672 }, { "epoch": 1.993766696349065, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.264521598815918, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8709348440170288, "num_tokens": 597886565.0, "step": 15673 }, { "epoch": 1.9938939066276555, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.243345260620117, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8735905885696411, "num_tokens": 597925097.0, "step": 15674 }, { "epoch": 1.994021116906246, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.365161895751953, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8836331367492676, "num_tokens": 597960786.0, "step": 15675 }, { "epoch": 1.9941483271848366, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.227801322937012, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8512696027755737, "num_tokens": 598000442.0, "step": 15676 }, { "epoch": 1.9942755374634271, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.340887069702148, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8678185343742371, "num_tokens": 598042946.0, "step": 15677 }, { "epoch": 1.9944027477420176, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.29144287109375, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8508294820785522, "num_tokens": 598081659.0, "step": 15678 }, { "epoch": 1.9945299580206082, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.348109245300293, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8651454448699951, "num_tokens": 598120203.0, "step": 15679 }, { "epoch": 1.9946571682991987, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.274506568908691, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.8574193716049194, "num_tokens": 598154779.0, "step": 15680 }, { "epoch": 1.9947843785777892, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.307063102722168, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8701622486114502, "num_tokens": 598197075.0, "step": 15681 }, { "epoch": 1.9949115888563798, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.25147533416748, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8627091646194458, "num_tokens": 598240539.0, "step": 15682 }, { "epoch": 1.99503879913497, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.337202072143555, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8719906806945801, "num_tokens": 598281776.0, "step": 15683 }, { "epoch": 1.9951660094135606, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.240799903869629, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8773375749588013, "num_tokens": 598322715.0, "step": 15684 }, { "epoch": 1.9952932196921511, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.330963134765625, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8649406433105469, "num_tokens": 598364495.0, "step": 15685 }, { "epoch": 1.9954204299707416, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.242122650146484, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8673491477966309, "num_tokens": 598407102.0, "step": 15686 }, { "epoch": 1.9955476402493322, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.311614990234375, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8574126362800598, "num_tokens": 598448600.0, "step": 15687 }, { "epoch": 1.9956748505279227, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.200726509094238, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8601524829864502, "num_tokens": 598491140.0, "step": 15688 }, { "epoch": 1.995802060806513, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.310420989990234, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8668193817138672, "num_tokens": 598527831.0, "step": 15689 }, { "epoch": 1.9959292710851035, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.265470504760742, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8574063777923584, "num_tokens": 598568680.0, "step": 15690 }, { "epoch": 1.996056481363694, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.326115608215332, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8590823411941528, "num_tokens": 598606254.0, "step": 15691 }, { "epoch": 1.9961836916422846, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.318229675292969, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.871322512626648, "num_tokens": 598642400.0, "step": 15692 }, { "epoch": 1.9963109019208751, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.268295288085938, "learning_rate": 1e-06, "loss": 0.4714, "mean_token_accuracy": 0.8482257127761841, "num_tokens": 598678379.0, "step": 15693 }, { "epoch": 1.9964381121994657, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.247139930725098, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8671408295631409, "num_tokens": 598718234.0, "step": 15694 }, { "epoch": 1.9965653224780562, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.284011840820312, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8727660179138184, "num_tokens": 598754897.0, "step": 15695 }, { "epoch": 1.9966925327566467, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.267882347106934, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8640846014022827, "num_tokens": 598785268.0, "step": 15696 }, { "epoch": 1.9968197430352372, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.24698543548584, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8761723041534424, "num_tokens": 598827131.0, "step": 15697 }, { "epoch": 1.9969469533138278, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.335290908813477, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.870455801486969, "num_tokens": 598859507.0, "step": 15698 }, { "epoch": 1.9970741635924183, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.251416206359863, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8605773448944092, "num_tokens": 598893280.0, "step": 15699 }, { "epoch": 1.9972013738710088, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.201911926269531, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.882630467414856, "num_tokens": 598932759.0, "step": 15700 }, { "epoch": 1.9973285841495994, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.349164962768555, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8800609111785889, "num_tokens": 598972899.0, "step": 15701 }, { "epoch": 1.9974557944281899, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.258901596069336, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8710371255874634, "num_tokens": 599012811.0, "step": 15702 }, { "epoch": 1.9975830047067804, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.325377464294434, "learning_rate": 1e-06, "loss": 0.4535, "mean_token_accuracy": 0.8608368635177612, "num_tokens": 599055632.0, "step": 15703 }, { "epoch": 1.997710214985371, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.2672758102417, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8521817922592163, "num_tokens": 599090924.0, "step": 15704 }, { "epoch": 1.9978374252639615, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.301168441772461, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8833504915237427, "num_tokens": 599130434.0, "step": 15705 }, { "epoch": 1.997964635542552, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.262537002563477, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8684644103050232, "num_tokens": 599171018.0, "step": 15706 }, { "epoch": 1.9980918458211423, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.350683212280273, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8614662885665894, "num_tokens": 599212445.0, "step": 15707 }, { "epoch": 1.9982190560997328, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.263472557067871, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8747676014900208, "num_tokens": 599250591.0, "step": 15708 }, { "epoch": 1.9983462663783234, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.328786849975586, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8653944730758667, "num_tokens": 599284987.0, "step": 15709 }, { "epoch": 1.9984734766569139, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.28913402557373, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8654906749725342, "num_tokens": 599324656.0, "step": 15710 }, { "epoch": 1.9986006869355044, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.313591003417969, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8801320791244507, "num_tokens": 599364014.0, "step": 15711 }, { "epoch": 1.998727897214095, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.348708152770996, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.861475944519043, "num_tokens": 599396781.0, "step": 15712 }, { "epoch": 1.9988551074926852, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.325465202331543, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8789451718330383, "num_tokens": 599433240.0, "step": 15713 }, { "epoch": 1.9989823177712758, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.270013809204102, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.878969669342041, "num_tokens": 599474883.0, "step": 15714 }, { "epoch": 1.9991095280498663, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.306077003479004, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8694987297058105, "num_tokens": 599508492.0, "step": 15715 }, { "epoch": 1.9992367383284568, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.318636894226074, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8718521595001221, "num_tokens": 599545946.0, "step": 15716 }, { "epoch": 1.9993639486070474, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.318487167358398, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8827985525131226, "num_tokens": 599586217.0, "step": 15717 }, { "epoch": 1.9994911588856379, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.27761459350586, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8767441511154175, "num_tokens": 599620431.0, "step": 15718 }, { "epoch": 1.9996183691642284, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.307896614074707, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8555111885070801, "num_tokens": 599659959.0, "step": 15719 }, { "epoch": 1.999745579442819, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.308525085449219, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8683243989944458, "num_tokens": 599696558.0, "step": 15720 }, { "epoch": 1.9998727897214095, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.28520393371582, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.858666181564331, "num_tokens": 599734925.0, "step": 15721 }, { "epoch": 2.0, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.31340217590332, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8736575841903687, "num_tokens": 599772613.0, "step": 15722 }, { "epoch": 2.0001272102785905, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.30822467803955, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8756750822067261, "num_tokens": 599813892.0, "step": 15723 }, { "epoch": 2.000254420557181, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.23826789855957, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.872077465057373, "num_tokens": 599852807.0, "step": 15724 }, { "epoch": 2.0003816308357716, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.340652465820312, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8788743019104004, "num_tokens": 599893712.0, "step": 15725 }, { "epoch": 2.000508841114362, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.316493034362793, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8740811347961426, "num_tokens": 599929954.0, "step": 15726 }, { "epoch": 2.0006360513929526, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.278362274169922, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8704235553741455, "num_tokens": 599967763.0, "step": 15727 }, { "epoch": 2.000763261671543, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.300101280212402, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8809567093849182, "num_tokens": 600008938.0, "step": 15728 }, { "epoch": 2.0008904719501337, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.310094833374023, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8689721822738647, "num_tokens": 600051849.0, "step": 15729 }, { "epoch": 2.0010176822287242, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.296900749206543, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8790825605392456, "num_tokens": 600088568.0, "step": 15730 }, { "epoch": 2.0011448925073148, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.318324089050293, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8769733905792236, "num_tokens": 600131553.0, "step": 15731 }, { "epoch": 2.0012721027859053, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.30329418182373, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8728113770484924, "num_tokens": 600173198.0, "step": 15732 }, { "epoch": 2.001399313064496, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.320962905883789, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8773059844970703, "num_tokens": 600211347.0, "step": 15733 }, { "epoch": 2.0015265233430863, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.340778350830078, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8661657571792603, "num_tokens": 600249953.0, "step": 15734 }, { "epoch": 2.0016537336216764, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.294808387756348, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8650180101394653, "num_tokens": 600291211.0, "step": 15735 }, { "epoch": 2.001780943900267, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.296012878417969, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8776916265487671, "num_tokens": 600332906.0, "step": 15736 }, { "epoch": 2.0019081541788575, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.336593627929688, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8729265928268433, "num_tokens": 600370570.0, "step": 15737 }, { "epoch": 2.002035364457448, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.287728309631348, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8632988929748535, "num_tokens": 600409368.0, "step": 15738 }, { "epoch": 2.0021625747360385, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.279340744018555, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8762176036834717, "num_tokens": 600442019.0, "step": 15739 }, { "epoch": 2.002289785014629, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.337063789367676, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8817691802978516, "num_tokens": 600471665.0, "step": 15740 }, { "epoch": 2.0024169952932196, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.350080490112305, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8711050748825073, "num_tokens": 600508804.0, "step": 15741 }, { "epoch": 2.00254420557181, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.354013442993164, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8612175583839417, "num_tokens": 600543895.0, "step": 15742 }, { "epoch": 2.0026714158504006, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.36628246307373, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8741426467895508, "num_tokens": 600585818.0, "step": 15743 }, { "epoch": 2.002798626128991, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.331117630004883, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8764944076538086, "num_tokens": 600623106.0, "step": 15744 }, { "epoch": 2.0029258364075817, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.27958869934082, "learning_rate": 1e-06, "loss": 0.33, "mean_token_accuracy": 0.8934279680252075, "num_tokens": 600662662.0, "step": 15745 }, { "epoch": 2.0030530466861722, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.328446388244629, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8596479892730713, "num_tokens": 600699416.0, "step": 15746 }, { "epoch": 2.0031802569647628, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.29789924621582, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8666819334030151, "num_tokens": 600739351.0, "step": 15747 }, { "epoch": 2.0033074672433533, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.253337860107422, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8753274083137512, "num_tokens": 600781198.0, "step": 15748 }, { "epoch": 2.003434677521944, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.30494213104248, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8713726997375488, "num_tokens": 600821743.0, "step": 15749 }, { "epoch": 2.0035618878005343, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.269244194030762, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8877013325691223, "num_tokens": 600862463.0, "step": 15750 }, { "epoch": 2.003689098079125, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.302125930786133, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8633027672767639, "num_tokens": 600902629.0, "step": 15751 }, { "epoch": 2.0038163083577154, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.323806762695312, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8714607954025269, "num_tokens": 600941778.0, "step": 15752 }, { "epoch": 2.003943518636306, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.278419494628906, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8810839653015137, "num_tokens": 600979425.0, "step": 15753 }, { "epoch": 2.0040707289148965, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.292211532592773, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8693457841873169, "num_tokens": 601017882.0, "step": 15754 }, { "epoch": 2.004197939193487, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.289809226989746, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8679051399230957, "num_tokens": 601053394.0, "step": 15755 }, { "epoch": 2.0043251494720775, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.24826717376709, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8766054511070251, "num_tokens": 601094940.0, "step": 15756 }, { "epoch": 2.004452359750668, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.323352813720703, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8736190795898438, "num_tokens": 601132963.0, "step": 15757 }, { "epoch": 2.0045795700292586, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.251151084899902, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8717470169067383, "num_tokens": 601168511.0, "step": 15758 }, { "epoch": 2.0047067803078487, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.293960571289062, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8732684850692749, "num_tokens": 601208922.0, "step": 15759 }, { "epoch": 2.004833990586439, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.283101081848145, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8668188452720642, "num_tokens": 601253814.0, "step": 15760 }, { "epoch": 2.0049612008650297, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.234573364257812, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8738375902175903, "num_tokens": 601295584.0, "step": 15761 }, { "epoch": 2.0050884111436202, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.29619312286377, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.865143895149231, "num_tokens": 601329134.0, "step": 15762 }, { "epoch": 2.0052156214222108, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.312695503234863, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8814501762390137, "num_tokens": 601369100.0, "step": 15763 }, { "epoch": 2.0053428317008013, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.282689094543457, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8877478837966919, "num_tokens": 601407876.0, "step": 15764 }, { "epoch": 2.005470041979392, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.250761985778809, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8856779932975769, "num_tokens": 601446996.0, "step": 15765 }, { "epoch": 2.0055972522579824, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.331433296203613, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8651350140571594, "num_tokens": 601482158.0, "step": 15766 }, { "epoch": 2.005724462536573, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.293730735778809, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8785737156867981, "num_tokens": 601517940.0, "step": 15767 }, { "epoch": 2.0058516728151634, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.270432472229004, "learning_rate": 1e-06, "loss": 0.3297, "mean_token_accuracy": 0.8932965993881226, "num_tokens": 601553823.0, "step": 15768 }, { "epoch": 2.005978883093754, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.30154037475586, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8612918853759766, "num_tokens": 601595482.0, "step": 15769 }, { "epoch": 2.0061060933723445, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.271018028259277, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8811586499214172, "num_tokens": 601632451.0, "step": 15770 }, { "epoch": 2.006233303650935, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.301124572753906, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8710683584213257, "num_tokens": 601667573.0, "step": 15771 }, { "epoch": 2.0063605139295255, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.305280685424805, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8790981769561768, "num_tokens": 601704494.0, "step": 15772 }, { "epoch": 2.006487724208116, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.26485824584961, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8638850450515747, "num_tokens": 601745307.0, "step": 15773 }, { "epoch": 2.0066149344867066, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.282938957214355, "learning_rate": 1e-06, "loss": 0.3503, "mean_token_accuracy": 0.8898051977157593, "num_tokens": 601782551.0, "step": 15774 }, { "epoch": 2.006742144765297, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.354402542114258, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.874950110912323, "num_tokens": 601823378.0, "step": 15775 }, { "epoch": 2.0068693550438876, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.30249309539795, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8736311793327332, "num_tokens": 601860881.0, "step": 15776 }, { "epoch": 2.006996565322478, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.243818283081055, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8721148371696472, "num_tokens": 601903502.0, "step": 15777 }, { "epoch": 2.0071237756010687, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.338713645935059, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.8909405469894409, "num_tokens": 601937097.0, "step": 15778 }, { "epoch": 2.007250985879659, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.239346504211426, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8773015737533569, "num_tokens": 601978050.0, "step": 15779 }, { "epoch": 2.0073781961582498, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.317418098449707, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8715768456459045, "num_tokens": 602018167.0, "step": 15780 }, { "epoch": 2.0075054064368403, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.273585319519043, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8587291836738586, "num_tokens": 602054485.0, "step": 15781 }, { "epoch": 2.007632616715431, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.3209867477417, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.879029393196106, "num_tokens": 602092881.0, "step": 15782 }, { "epoch": 2.0077598269940213, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.7896089553833, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8750027418136597, "num_tokens": 602127424.0, "step": 15783 }, { "epoch": 2.0078870372726114, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.27900218963623, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8805757761001587, "num_tokens": 602168340.0, "step": 15784 }, { "epoch": 2.008014247551202, "ewc_loss": 0.03369140625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 13.196313858032227, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8636425733566284, "num_tokens": 602207894.0, "step": 15785 }, { "epoch": 2.0081414578297925, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.346078872680664, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8756123781204224, "num_tokens": 602245924.0, "step": 15786 }, { "epoch": 2.008268668108383, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.256331443786621, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8720447421073914, "num_tokens": 602278417.0, "step": 15787 }, { "epoch": 2.0083958783869735, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.22957706451416, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8761276602745056, "num_tokens": 602318431.0, "step": 15788 }, { "epoch": 2.008523088665564, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.313872337341309, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8635116219520569, "num_tokens": 602356315.0, "step": 15789 }, { "epoch": 2.0086502989441546, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.279391288757324, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.8879390954971313, "num_tokens": 602398865.0, "step": 15790 }, { "epoch": 2.008777509222745, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.240068435668945, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8837764263153076, "num_tokens": 602435835.0, "step": 15791 }, { "epoch": 2.0089047195013356, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.290843963623047, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8822291493415833, "num_tokens": 602470359.0, "step": 15792 }, { "epoch": 2.009031929779926, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.234288215637207, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8774935007095337, "num_tokens": 602512132.0, "step": 15793 }, { "epoch": 2.0091591400585167, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.282268524169922, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8703207969665527, "num_tokens": 602552700.0, "step": 15794 }, { "epoch": 2.0092863503371072, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.287609100341797, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8741374015808105, "num_tokens": 602587779.0, "step": 15795 }, { "epoch": 2.0094135606156978, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.271875381469727, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8642950057983398, "num_tokens": 602623834.0, "step": 15796 }, { "epoch": 2.0095407708942883, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.320241928100586, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8844514489173889, "num_tokens": 602660141.0, "step": 15797 }, { "epoch": 2.009667981172879, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3602933883667, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8773279190063477, "num_tokens": 602698990.0, "step": 15798 }, { "epoch": 2.0097951914514693, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.347763061523438, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8565245866775513, "num_tokens": 602732053.0, "step": 15799 }, { "epoch": 2.00992240173006, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.261712074279785, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8748781681060791, "num_tokens": 602770975.0, "step": 15800 }, { "epoch": 2.0100496120086504, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.338151931762695, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8652437329292297, "num_tokens": 602813207.0, "step": 15801 }, { "epoch": 2.010176822287241, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.310111999511719, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8793503642082214, "num_tokens": 602849951.0, "step": 15802 }, { "epoch": 2.0103040325658315, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.298221588134766, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8643961548805237, "num_tokens": 602891418.0, "step": 15803 }, { "epoch": 2.010431242844422, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.32802963256836, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8799540996551514, "num_tokens": 602933039.0, "step": 15804 }, { "epoch": 2.0105584531230125, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.337430000305176, "learning_rate": 1e-06, "loss": 0.3545, "mean_token_accuracy": 0.8886302709579468, "num_tokens": 602965546.0, "step": 15805 }, { "epoch": 2.010685663401603, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.32383918762207, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8735733032226562, "num_tokens": 603004473.0, "step": 15806 }, { "epoch": 2.0108128736801936, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.337992668151855, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8669044971466064, "num_tokens": 603041644.0, "step": 15807 }, { "epoch": 2.0109400839587837, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.308034896850586, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.862682044506073, "num_tokens": 603084698.0, "step": 15808 }, { "epoch": 2.011067294237374, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.348992347717285, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8587433099746704, "num_tokens": 603121059.0, "step": 15809 }, { "epoch": 2.0111945045159647, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.364234924316406, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8717313408851624, "num_tokens": 603156604.0, "step": 15810 }, { "epoch": 2.0113217147945552, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.335744857788086, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8739945888519287, "num_tokens": 603183743.0, "step": 15811 }, { "epoch": 2.0114489250731458, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.33011245727539, "learning_rate": 1e-06, "loss": 0.4721, "mean_token_accuracy": 0.8504857420921326, "num_tokens": 603222099.0, "step": 15812 }, { "epoch": 2.0115761353517363, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.385600090026855, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8549257516860962, "num_tokens": 603263467.0, "step": 15813 }, { "epoch": 2.011703345630327, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.300135612487793, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8552664518356323, "num_tokens": 603296726.0, "step": 15814 }, { "epoch": 2.0118305559089174, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.38140869140625, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8792740106582642, "num_tokens": 603329691.0, "step": 15815 }, { "epoch": 2.011957766187508, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.257219314575195, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8813434839248657, "num_tokens": 603371619.0, "step": 15816 }, { "epoch": 2.0120849764660984, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.316709518432617, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8750388622283936, "num_tokens": 603400018.0, "step": 15817 }, { "epoch": 2.012212186744689, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.302254676818848, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8769288063049316, "num_tokens": 603436971.0, "step": 15818 }, { "epoch": 2.0123393970232795, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.35031795501709, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8739117383956909, "num_tokens": 603472802.0, "step": 15819 }, { "epoch": 2.01246660730187, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3164644241333, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8796733617782593, "num_tokens": 603506547.0, "step": 15820 }, { "epoch": 2.0125938175804605, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.316051483154297, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8577094674110413, "num_tokens": 603544967.0, "step": 15821 }, { "epoch": 2.012721027859051, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.322169303894043, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8683749437332153, "num_tokens": 603582522.0, "step": 15822 }, { "epoch": 2.0128482381376416, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.315999031066895, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8828948736190796, "num_tokens": 603619584.0, "step": 15823 }, { "epoch": 2.012975448416232, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.322295188903809, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8614211082458496, "num_tokens": 603660542.0, "step": 15824 }, { "epoch": 2.0131026586948226, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.35280990600586, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8748464584350586, "num_tokens": 603698900.0, "step": 15825 }, { "epoch": 2.013229868973413, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.344152450561523, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8565616011619568, "num_tokens": 603733019.0, "step": 15826 }, { "epoch": 2.0133570792520037, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.29293155670166, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8818467259407043, "num_tokens": 603772071.0, "step": 15827 }, { "epoch": 2.013484289530594, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.32284927368164, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8657432794570923, "num_tokens": 603811008.0, "step": 15828 }, { "epoch": 2.0136114998091847, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.322128295898438, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8636472225189209, "num_tokens": 603852647.0, "step": 15829 }, { "epoch": 2.0137387100877753, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.3484468460083, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8721323013305664, "num_tokens": 603894198.0, "step": 15830 }, { "epoch": 2.013865920366366, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.346349716186523, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8700959086418152, "num_tokens": 603930168.0, "step": 15831 }, { "epoch": 2.0139931306449563, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.351316452026367, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8829399347305298, "num_tokens": 603966189.0, "step": 15832 }, { "epoch": 2.0141203409235464, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.334031105041504, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8696666955947876, "num_tokens": 604011110.0, "step": 15833 }, { "epoch": 2.014247551202137, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.369485855102539, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8561100959777832, "num_tokens": 604044256.0, "step": 15834 }, { "epoch": 2.0143747614807275, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.352367401123047, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8820597529411316, "num_tokens": 604078652.0, "step": 15835 }, { "epoch": 2.014501971759318, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.311281204223633, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8819465637207031, "num_tokens": 604112302.0, "step": 15836 }, { "epoch": 2.0146291820379085, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.349964141845703, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8883670568466187, "num_tokens": 604150011.0, "step": 15837 }, { "epoch": 2.014756392316499, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.329975128173828, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8730781078338623, "num_tokens": 604188859.0, "step": 15838 }, { "epoch": 2.0148836025950896, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.326927185058594, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8682957291603088, "num_tokens": 604232273.0, "step": 15839 }, { "epoch": 2.01501081287368, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.32557487487793, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8804726004600525, "num_tokens": 604265236.0, "step": 15840 }, { "epoch": 2.0151380231522706, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.276779174804688, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8867889642715454, "num_tokens": 604304784.0, "step": 15841 }, { "epoch": 2.015265233430861, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.392147064208984, "learning_rate": 1e-06, "loss": 0.3207, "mean_token_accuracy": 0.898457407951355, "num_tokens": 604341461.0, "step": 15842 }, { "epoch": 2.0153924437094517, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.325067520141602, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.879310131072998, "num_tokens": 604378628.0, "step": 15843 }, { "epoch": 2.0155196539880422, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.323226928710938, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8830864429473877, "num_tokens": 604418070.0, "step": 15844 }, { "epoch": 2.0156468642666328, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.302494049072266, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8859513998031616, "num_tokens": 604454445.0, "step": 15845 }, { "epoch": 2.0157740745452233, "ewc_loss": 0.033935546875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 13.296408653259277, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8774452209472656, "num_tokens": 604488559.0, "step": 15846 }, { "epoch": 2.015901284823814, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.348981857299805, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8661656379699707, "num_tokens": 604522538.0, "step": 15847 }, { "epoch": 2.0160284951024043, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.27853012084961, "learning_rate": 1e-06, "loss": 0.3293, "mean_token_accuracy": 0.89216148853302, "num_tokens": 604561750.0, "step": 15848 }, { "epoch": 2.016155705380995, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.313298225402832, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8874676823616028, "num_tokens": 604594961.0, "step": 15849 }, { "epoch": 2.0162829156595854, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.259570121765137, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8649481534957886, "num_tokens": 604631301.0, "step": 15850 }, { "epoch": 2.016410125938176, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.372750282287598, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8542764186859131, "num_tokens": 604666049.0, "step": 15851 }, { "epoch": 2.0165373362167665, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.293886184692383, "learning_rate": 1e-06, "loss": 0.334, "mean_token_accuracy": 0.8944398760795593, "num_tokens": 604701984.0, "step": 15852 }, { "epoch": 2.016664546495357, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.276308059692383, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8821337819099426, "num_tokens": 604738414.0, "step": 15853 }, { "epoch": 2.0167917567739475, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.298039436340332, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8760140538215637, "num_tokens": 604779961.0, "step": 15854 }, { "epoch": 2.016918967052538, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.316311836242676, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8676391839981079, "num_tokens": 604819729.0, "step": 15855 }, { "epoch": 2.0170461773311286, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.343771934509277, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.883558452129364, "num_tokens": 604864108.0, "step": 15856 }, { "epoch": 2.0171733876097186, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.270381927490234, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8684525489807129, "num_tokens": 604905582.0, "step": 15857 }, { "epoch": 2.017300597888309, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.28684139251709, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8817567825317383, "num_tokens": 604943931.0, "step": 15858 }, { "epoch": 2.0174278081668997, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.334324836730957, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.869691014289856, "num_tokens": 604982100.0, "step": 15859 }, { "epoch": 2.0175550184454902, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.348387718200684, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8627110719680786, "num_tokens": 605026289.0, "step": 15860 }, { "epoch": 2.0176822287240808, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.292938232421875, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8777303695678711, "num_tokens": 605070214.0, "step": 15861 }, { "epoch": 2.0178094390026713, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.326617240905762, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8695274591445923, "num_tokens": 605113495.0, "step": 15862 }, { "epoch": 2.017936649281262, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.323080062866211, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8782461285591125, "num_tokens": 605145717.0, "step": 15863 }, { "epoch": 2.0180638595598523, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.350380897521973, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8672178983688354, "num_tokens": 605180930.0, "step": 15864 }, { "epoch": 2.018191069838443, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.335638046264648, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8804078698158264, "num_tokens": 605214004.0, "step": 15865 }, { "epoch": 2.0183182801170334, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.308758735656738, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8743806481361389, "num_tokens": 605253244.0, "step": 15866 }, { "epoch": 2.018445490395624, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.333178520202637, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8688006401062012, "num_tokens": 605293832.0, "step": 15867 }, { "epoch": 2.0185727006742145, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.333918571472168, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8847982287406921, "num_tokens": 605333811.0, "step": 15868 }, { "epoch": 2.018699910952805, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.31606388092041, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8744014501571655, "num_tokens": 605371560.0, "step": 15869 }, { "epoch": 2.0188271212313955, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.289835929870605, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8713757991790771, "num_tokens": 605410555.0, "step": 15870 }, { "epoch": 2.018954331509986, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.318316459655762, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8680684566497803, "num_tokens": 605451851.0, "step": 15871 }, { "epoch": 2.0190815417885766, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.301839828491211, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8659717440605164, "num_tokens": 605490442.0, "step": 15872 }, { "epoch": 2.019208752067167, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.373748779296875, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8729385137557983, "num_tokens": 605525237.0, "step": 15873 }, { "epoch": 2.0193359623457576, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.340374946594238, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8791411519050598, "num_tokens": 605563624.0, "step": 15874 }, { "epoch": 2.019463172624348, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.391438484191895, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8833965063095093, "num_tokens": 605597419.0, "step": 15875 }, { "epoch": 2.0195903829029387, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.393970489501953, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.867218554019928, "num_tokens": 605639954.0, "step": 15876 }, { "epoch": 2.019717593181529, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.31470775604248, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8746187686920166, "num_tokens": 605678226.0, "step": 15877 }, { "epoch": 2.0198448034601197, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.37757396697998, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8825281262397766, "num_tokens": 605719925.0, "step": 15878 }, { "epoch": 2.0199720137387103, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.336398124694824, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8545145988464355, "num_tokens": 605761282.0, "step": 15879 }, { "epoch": 2.020099224017301, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.393514633178711, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8830781579017639, "num_tokens": 605800681.0, "step": 15880 }, { "epoch": 2.0202264342958913, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.356081008911133, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8571122884750366, "num_tokens": 605836282.0, "step": 15881 }, { "epoch": 2.0203536445744814, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.330353736877441, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8760414123535156, "num_tokens": 605880360.0, "step": 15882 }, { "epoch": 2.020480854853072, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.331746101379395, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.8839316964149475, "num_tokens": 605918363.0, "step": 15883 }, { "epoch": 2.0206080651316625, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.354307174682617, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8874229788780212, "num_tokens": 605953032.0, "step": 15884 }, { "epoch": 2.020735275410253, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.327526092529297, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8811867833137512, "num_tokens": 605992633.0, "step": 15885 }, { "epoch": 2.0208624856888435, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.35489273071289, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8759125471115112, "num_tokens": 606031776.0, "step": 15886 }, { "epoch": 2.020989695967434, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.288166046142578, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.868199348449707, "num_tokens": 606070614.0, "step": 15887 }, { "epoch": 2.0211169062460246, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.361926078796387, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.873536229133606, "num_tokens": 606115898.0, "step": 15888 }, { "epoch": 2.021244116524615, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.369792938232422, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8780355453491211, "num_tokens": 606150231.0, "step": 15889 }, { "epoch": 2.0213713268032056, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.29253101348877, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8657954931259155, "num_tokens": 606191527.0, "step": 15890 }, { "epoch": 2.021498537081796, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.399815559387207, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8669363260269165, "num_tokens": 606229070.0, "step": 15891 }, { "epoch": 2.0216257473603867, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.268378257751465, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8750547170639038, "num_tokens": 606269555.0, "step": 15892 }, { "epoch": 2.021752957638977, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.392892837524414, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8688774704933167, "num_tokens": 606302040.0, "step": 15893 }, { "epoch": 2.0218801679175677, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.261808395385742, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8833663463592529, "num_tokens": 606341769.0, "step": 15894 }, { "epoch": 2.0220073781961583, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.342812538146973, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8613334894180298, "num_tokens": 606379989.0, "step": 15895 }, { "epoch": 2.022134588474749, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.328423500061035, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8687999248504639, "num_tokens": 606414396.0, "step": 15896 }, { "epoch": 2.0222617987533393, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.388707160949707, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8771266937255859, "num_tokens": 606445949.0, "step": 15897 }, { "epoch": 2.02238900903193, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.360353469848633, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.876228928565979, "num_tokens": 606484751.0, "step": 15898 }, { "epoch": 2.0225162193105204, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.30687141418457, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8767843246459961, "num_tokens": 606525390.0, "step": 15899 }, { "epoch": 2.022643429589111, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.350728988647461, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8702288269996643, "num_tokens": 606563364.0, "step": 15900 }, { "epoch": 2.0227706398677014, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.329935073852539, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8769209980964661, "num_tokens": 606598140.0, "step": 15901 }, { "epoch": 2.022897850146292, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.359994888305664, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.866716206073761, "num_tokens": 606636480.0, "step": 15902 }, { "epoch": 2.0230250604248825, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.266852378845215, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8776289820671082, "num_tokens": 606676044.0, "step": 15903 }, { "epoch": 2.023152270703473, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.382433891296387, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8675136566162109, "num_tokens": 606713262.0, "step": 15904 }, { "epoch": 2.0232794809820636, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.2636137008667, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8711782693862915, "num_tokens": 606754844.0, "step": 15905 }, { "epoch": 2.0234066912606536, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.334243774414062, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.873888373374939, "num_tokens": 606800910.0, "step": 15906 }, { "epoch": 2.023533901539244, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.306353569030762, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8695621490478516, "num_tokens": 606841855.0, "step": 15907 }, { "epoch": 2.0236611118178347, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.298344612121582, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8633672595024109, "num_tokens": 606888261.0, "step": 15908 }, { "epoch": 2.0237883220964252, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.335007667541504, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.864474356174469, "num_tokens": 606926235.0, "step": 15909 }, { "epoch": 2.0239155323750158, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.38725757598877, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8699367046356201, "num_tokens": 606962527.0, "step": 15910 }, { "epoch": 2.0240427426536063, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.321958541870117, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8784065246582031, "num_tokens": 607008924.0, "step": 15911 }, { "epoch": 2.024169952932197, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.352860450744629, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8702137470245361, "num_tokens": 607049788.0, "step": 15912 }, { "epoch": 2.0242971632107873, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.342987060546875, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8669246435165405, "num_tokens": 607087592.0, "step": 15913 }, { "epoch": 2.024424373489378, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.377930641174316, "learning_rate": 1e-06, "loss": 0.3436, "mean_token_accuracy": 0.8889659643173218, "num_tokens": 607119954.0, "step": 15914 }, { "epoch": 2.0245515837679684, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.346826553344727, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8643515706062317, "num_tokens": 607156511.0, "step": 15915 }, { "epoch": 2.024678794046559, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.357905387878418, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8831027746200562, "num_tokens": 607198292.0, "step": 15916 }, { "epoch": 2.0248060043251495, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.37964916229248, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8714423775672913, "num_tokens": 607237572.0, "step": 15917 }, { "epoch": 2.02493321460374, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.415702819824219, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8641716837882996, "num_tokens": 607282321.0, "step": 15918 }, { "epoch": 2.0250604248823305, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.367155075073242, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8723820447921753, "num_tokens": 607313355.0, "step": 15919 }, { "epoch": 2.025187635160921, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.38965892791748, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8711779117584229, "num_tokens": 607351961.0, "step": 15920 }, { "epoch": 2.0253148454395116, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.340805053710938, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8756325840950012, "num_tokens": 607391368.0, "step": 15921 }, { "epoch": 2.025442055718102, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.300070762634277, "learning_rate": 1e-06, "loss": 0.3366, "mean_token_accuracy": 0.8918723464012146, "num_tokens": 607427328.0, "step": 15922 }, { "epoch": 2.0255692659966926, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.333213806152344, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8697584271430969, "num_tokens": 607468243.0, "step": 15923 }, { "epoch": 2.025696476275283, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.38382339477539, "learning_rate": 1e-06, "loss": 0.4539, "mean_token_accuracy": 0.8574464321136475, "num_tokens": 607502466.0, "step": 15924 }, { "epoch": 2.0258236865538737, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.33522891998291, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.865248441696167, "num_tokens": 607544093.0, "step": 15925 }, { "epoch": 2.025950896832464, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.367679595947266, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8803556561470032, "num_tokens": 607576054.0, "step": 15926 }, { "epoch": 2.0260781071110547, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.337810516357422, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8860963582992554, "num_tokens": 607608462.0, "step": 15927 }, { "epoch": 2.0262053173896453, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.37263011932373, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8681890964508057, "num_tokens": 607645272.0, "step": 15928 }, { "epoch": 2.026332527668236, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.355835914611816, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8764723539352417, "num_tokens": 607682585.0, "step": 15929 }, { "epoch": 2.0264597379468263, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.409131050109863, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8907655477523804, "num_tokens": 607721302.0, "step": 15930 }, { "epoch": 2.0265869482254164, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.321526527404785, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8693300485610962, "num_tokens": 607758357.0, "step": 15931 }, { "epoch": 2.026714158504007, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.338658332824707, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8743882179260254, "num_tokens": 607795467.0, "step": 15932 }, { "epoch": 2.0268413687825975, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.326285362243652, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8687852621078491, "num_tokens": 607833192.0, "step": 15933 }, { "epoch": 2.026968579061188, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.345149993896484, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8828996419906616, "num_tokens": 607866672.0, "step": 15934 }, { "epoch": 2.0270957893397785, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.351157188415527, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8746529221534729, "num_tokens": 607907976.0, "step": 15935 }, { "epoch": 2.027222999618369, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.297801971435547, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8770452737808228, "num_tokens": 607946009.0, "step": 15936 }, { "epoch": 2.0273502098969596, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.312499046325684, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8788291811943054, "num_tokens": 607986800.0, "step": 15937 }, { "epoch": 2.02747742017555, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.340974807739258, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8678147792816162, "num_tokens": 608021807.0, "step": 15938 }, { "epoch": 2.0276046304541406, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.284208297729492, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8718200922012329, "num_tokens": 608060294.0, "step": 15939 }, { "epoch": 2.027731840732731, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.37837028503418, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.872679591178894, "num_tokens": 608100790.0, "step": 15940 }, { "epoch": 2.0278590510113217, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.277718544006348, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8699914216995239, "num_tokens": 608142164.0, "step": 15941 }, { "epoch": 2.027986261289912, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.337560653686523, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8745647072792053, "num_tokens": 608184175.0, "step": 15942 }, { "epoch": 2.0281134715685027, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.306140899658203, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8682129383087158, "num_tokens": 608226801.0, "step": 15943 }, { "epoch": 2.0282406818470933, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.392523765563965, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8819913864135742, "num_tokens": 608265155.0, "step": 15944 }, { "epoch": 2.028367892125684, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.326923370361328, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8843010663986206, "num_tokens": 608303817.0, "step": 15945 }, { "epoch": 2.0284951024042743, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.351099967956543, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8801429271697998, "num_tokens": 608338845.0, "step": 15946 }, { "epoch": 2.028622312682865, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.340497016906738, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8732882142066956, "num_tokens": 608377661.0, "step": 15947 }, { "epoch": 2.0287495229614554, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.38155460357666, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8736629486083984, "num_tokens": 608412897.0, "step": 15948 }, { "epoch": 2.028876733240046, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.325309753417969, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8651953935623169, "num_tokens": 608453584.0, "step": 15949 }, { "epoch": 2.0290039435186364, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.387163162231445, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.870455265045166, "num_tokens": 608492211.0, "step": 15950 }, { "epoch": 2.029131153797227, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.337793350219727, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.870220422744751, "num_tokens": 608534317.0, "step": 15951 }, { "epoch": 2.0292583640758175, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.337894439697266, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8828431367874146, "num_tokens": 608569763.0, "step": 15952 }, { "epoch": 2.029385574354408, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.355591773986816, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8770240545272827, "num_tokens": 608612145.0, "step": 15953 }, { "epoch": 2.0295127846329986, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.361701011657715, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8732188940048218, "num_tokens": 608649528.0, "step": 15954 }, { "epoch": 2.0296399949115886, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.380001068115234, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8673384189605713, "num_tokens": 608686973.0, "step": 15955 }, { "epoch": 2.029767205190179, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.336413383483887, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8757166862487793, "num_tokens": 608729169.0, "step": 15956 }, { "epoch": 2.0298944154687697, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.384138107299805, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8726142644882202, "num_tokens": 608764502.0, "step": 15957 }, { "epoch": 2.0300216257473602, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.369697570800781, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8743487000465393, "num_tokens": 608805190.0, "step": 15958 }, { "epoch": 2.0301488360259508, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.4033842086792, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8803956508636475, "num_tokens": 608840006.0, "step": 15959 }, { "epoch": 2.0302760463045413, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.356889724731445, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8815919160842896, "num_tokens": 608871551.0, "step": 15960 }, { "epoch": 2.030403256583132, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.36143970489502, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8527736067771912, "num_tokens": 608909610.0, "step": 15961 }, { "epoch": 2.0305304668617223, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.403517723083496, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8701568245887756, "num_tokens": 608950758.0, "step": 15962 }, { "epoch": 2.030657677140313, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.312352180480957, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8611814975738525, "num_tokens": 608990695.0, "step": 15963 }, { "epoch": 2.0307848874189034, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.368892669677734, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8878275156021118, "num_tokens": 609030978.0, "step": 15964 }, { "epoch": 2.030912097697494, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.323370933532715, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.88335782289505, "num_tokens": 609071974.0, "step": 15965 }, { "epoch": 2.0310393079760845, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.381117820739746, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8763487339019775, "num_tokens": 609111268.0, "step": 15966 }, { "epoch": 2.031166518254675, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.402254104614258, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.869143009185791, "num_tokens": 609149681.0, "step": 15967 }, { "epoch": 2.0312937285332655, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.35728931427002, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8791496157646179, "num_tokens": 609189121.0, "step": 15968 }, { "epoch": 2.031420938811856, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.36662769317627, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8829900026321411, "num_tokens": 609228070.0, "step": 15969 }, { "epoch": 2.0315481490904466, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.349194526672363, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8764564990997314, "num_tokens": 609269718.0, "step": 15970 }, { "epoch": 2.031675359369037, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.393498420715332, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8770724534988403, "num_tokens": 609307236.0, "step": 15971 }, { "epoch": 2.0318025696476276, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.342497825622559, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8561269044876099, "num_tokens": 609342993.0, "step": 15972 }, { "epoch": 2.031929779926218, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.425886154174805, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8798791766166687, "num_tokens": 609376817.0, "step": 15973 }, { "epoch": 2.0320569902048087, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.347909927368164, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8724828958511353, "num_tokens": 609411209.0, "step": 15974 }, { "epoch": 2.032184200483399, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.37344741821289, "learning_rate": 1e-06, "loss": 0.3398, "mean_token_accuracy": 0.8914629220962524, "num_tokens": 609449055.0, "step": 15975 }, { "epoch": 2.0323114107619897, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3522310256958, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8746504783630371, "num_tokens": 609486332.0, "step": 15976 }, { "epoch": 2.0324386210405803, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.328551292419434, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8607496023178101, "num_tokens": 609526924.0, "step": 15977 }, { "epoch": 2.032565831319171, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.327558517456055, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8879501819610596, "num_tokens": 609559542.0, "step": 15978 }, { "epoch": 2.032693041597761, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.337081909179688, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8670861721038818, "num_tokens": 609598989.0, "step": 15979 }, { "epoch": 2.0328202518763514, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.373886108398438, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8621362447738647, "num_tokens": 609635790.0, "step": 15980 }, { "epoch": 2.032947462154942, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.319931983947754, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8663691282272339, "num_tokens": 609670120.0, "step": 15981 }, { "epoch": 2.0330746724335325, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.29508113861084, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8733004331588745, "num_tokens": 609713260.0, "step": 15982 }, { "epoch": 2.033201882712123, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.350982666015625, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8808211088180542, "num_tokens": 609753236.0, "step": 15983 }, { "epoch": 2.0333290929907135, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3643159866333, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8772965669631958, "num_tokens": 609791629.0, "step": 15984 }, { "epoch": 2.033456303269304, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.307982444763184, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8586198091506958, "num_tokens": 609833457.0, "step": 15985 }, { "epoch": 2.0335835135478946, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.315181732177734, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8666259050369263, "num_tokens": 609866007.0, "step": 15986 }, { "epoch": 2.033710723826485, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.395630836486816, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8779454231262207, "num_tokens": 609903346.0, "step": 15987 }, { "epoch": 2.0338379341050756, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.397761344909668, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8776941895484924, "num_tokens": 609935885.0, "step": 15988 }, { "epoch": 2.033965144383666, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.410733222961426, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8676687479019165, "num_tokens": 609971127.0, "step": 15989 }, { "epoch": 2.0340923546622567, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.349596977233887, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8655454516410828, "num_tokens": 610016719.0, "step": 15990 }, { "epoch": 2.034219564940847, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.369215965270996, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8633582592010498, "num_tokens": 610057002.0, "step": 15991 }, { "epoch": 2.0343467752194377, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.546688079833984, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8789835572242737, "num_tokens": 610092779.0, "step": 15992 }, { "epoch": 2.0344739854980283, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.3411283493042, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8643447756767273, "num_tokens": 610137092.0, "step": 15993 }, { "epoch": 2.034601195776619, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.374436378479004, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8834104537963867, "num_tokens": 610178632.0, "step": 15994 }, { "epoch": 2.0347284060552093, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.372977256774902, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8722448945045471, "num_tokens": 610219789.0, "step": 15995 }, { "epoch": 2.0348556163338, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.365938186645508, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8596034049987793, "num_tokens": 610259104.0, "step": 15996 }, { "epoch": 2.0349828266123904, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.360029220581055, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8671355247497559, "num_tokens": 610296317.0, "step": 15997 }, { "epoch": 2.035110036890981, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.282769203186035, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8715702295303345, "num_tokens": 610339300.0, "step": 15998 }, { "epoch": 2.0352372471695714, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.390277862548828, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.869539737701416, "num_tokens": 610373917.0, "step": 15999 }, { "epoch": 2.035364457448162, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.342717170715332, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8727205395698547, "num_tokens": 610419571.0, "step": 16000 }, { "epoch": 2.0354916677267525, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.335667610168457, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8649352192878723, "num_tokens": 610458073.0, "step": 16001 }, { "epoch": 2.035618878005343, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.370731353759766, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8633810877799988, "num_tokens": 610497223.0, "step": 16002 }, { "epoch": 2.0357460882839336, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.36585521697998, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8536330461502075, "num_tokens": 610532563.0, "step": 16003 }, { "epoch": 2.0358732985625236, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.340971946716309, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8845077157020569, "num_tokens": 610566644.0, "step": 16004 }, { "epoch": 2.036000508841114, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.384682655334473, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8606016635894775, "num_tokens": 610604755.0, "step": 16005 }, { "epoch": 2.0361277191197047, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.324163436889648, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8778378963470459, "num_tokens": 610645905.0, "step": 16006 }, { "epoch": 2.036254929398295, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.396480560302734, "learning_rate": 1e-06, "loss": 0.3463, "mean_token_accuracy": 0.889723539352417, "num_tokens": 610686800.0, "step": 16007 }, { "epoch": 2.0363821396768857, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.365509033203125, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.874181866645813, "num_tokens": 610728291.0, "step": 16008 }, { "epoch": 2.0365093499554763, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.44445514678955, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8716608285903931, "num_tokens": 610759818.0, "step": 16009 }, { "epoch": 2.036636560234067, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.307294845581055, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8600559830665588, "num_tokens": 610800125.0, "step": 16010 }, { "epoch": 2.0367637705126573, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.364274978637695, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8638313412666321, "num_tokens": 610839261.0, "step": 16011 }, { "epoch": 2.036890980791248, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.371233940124512, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8780191540718079, "num_tokens": 610879092.0, "step": 16012 }, { "epoch": 2.0370181910698384, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.36014175415039, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8736078143119812, "num_tokens": 610914494.0, "step": 16013 }, { "epoch": 2.037145401348429, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.382741928100586, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8828698396682739, "num_tokens": 610950784.0, "step": 16014 }, { "epoch": 2.0372726116270194, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.395899772644043, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.863115668296814, "num_tokens": 610989254.0, "step": 16015 }, { "epoch": 2.03739982190561, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.350258827209473, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8657225370407104, "num_tokens": 611022668.0, "step": 16016 }, { "epoch": 2.0375270321842005, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.39188003540039, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.872532844543457, "num_tokens": 611062110.0, "step": 16017 }, { "epoch": 2.037654242462791, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.341567039489746, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8782123327255249, "num_tokens": 611098663.0, "step": 16018 }, { "epoch": 2.0377814527413816, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.364391326904297, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8698009252548218, "num_tokens": 611134058.0, "step": 16019 }, { "epoch": 2.037908663019972, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.320744514465332, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8740221858024597, "num_tokens": 611174748.0, "step": 16020 }, { "epoch": 2.0380358732985626, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.360280990600586, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.889218807220459, "num_tokens": 611216792.0, "step": 16021 }, { "epoch": 2.038163083577153, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.319857597351074, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8831512928009033, "num_tokens": 611253996.0, "step": 16022 }, { "epoch": 2.0382902938557437, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.386106491088867, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8839421272277832, "num_tokens": 611295351.0, "step": 16023 }, { "epoch": 2.038417504134334, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.353414535522461, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.872840940952301, "num_tokens": 611330513.0, "step": 16024 }, { "epoch": 2.0385447144129247, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.395025253295898, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8816133737564087, "num_tokens": 611369679.0, "step": 16025 }, { "epoch": 2.0386719246915153, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.354655265808105, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8806686401367188, "num_tokens": 611410398.0, "step": 16026 }, { "epoch": 2.038799134970106, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.333946228027344, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8674282431602478, "num_tokens": 611447151.0, "step": 16027 }, { "epoch": 2.0389263452486963, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.31507682800293, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8651660084724426, "num_tokens": 611486923.0, "step": 16028 }, { "epoch": 2.0390535555272864, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.376526832580566, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8802779316902161, "num_tokens": 611519502.0, "step": 16029 }, { "epoch": 2.039180765805877, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.332043647766113, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8598750829696655, "num_tokens": 611557647.0, "step": 16030 }, { "epoch": 2.0393079760844675, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.323784828186035, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8714390397071838, "num_tokens": 611597405.0, "step": 16031 }, { "epoch": 2.039435186363058, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.379779815673828, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8674277663230896, "num_tokens": 611635996.0, "step": 16032 }, { "epoch": 2.0395623966416485, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.326945304870605, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8812397122383118, "num_tokens": 611677900.0, "step": 16033 }, { "epoch": 2.039689606920239, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.352712631225586, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8765919804573059, "num_tokens": 611719184.0, "step": 16034 }, { "epoch": 2.0398168171988296, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.405754089355469, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8720966577529907, "num_tokens": 611761270.0, "step": 16035 }, { "epoch": 2.03994402747742, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.307785987854004, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8725634217262268, "num_tokens": 611794939.0, "step": 16036 }, { "epoch": 2.0400712377560106, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.398338317871094, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8840740919113159, "num_tokens": 611834093.0, "step": 16037 }, { "epoch": 2.040198448034601, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.351922035217285, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8684319257736206, "num_tokens": 611871327.0, "step": 16038 }, { "epoch": 2.0403256583131917, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.408173561096191, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.86543869972229, "num_tokens": 611907837.0, "step": 16039 }, { "epoch": 2.040452868591782, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.319210052490234, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8548887968063354, "num_tokens": 611950964.0, "step": 16040 }, { "epoch": 2.0405800788703727, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3938627243042, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8687087297439575, "num_tokens": 611985867.0, "step": 16041 }, { "epoch": 2.0407072891489633, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.4041109085083, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8766990900039673, "num_tokens": 612017819.0, "step": 16042 }, { "epoch": 2.040834499427554, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.389507293701172, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8588073253631592, "num_tokens": 612060669.0, "step": 16043 }, { "epoch": 2.0409617097061443, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.378746032714844, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8593665361404419, "num_tokens": 612099801.0, "step": 16044 }, { "epoch": 2.041088919984735, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.350442886352539, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8773211240768433, "num_tokens": 612139454.0, "step": 16045 }, { "epoch": 2.0412161302633254, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.396629333496094, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.853743314743042, "num_tokens": 612181695.0, "step": 16046 }, { "epoch": 2.041343340541916, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.31425952911377, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8874841928482056, "num_tokens": 612223200.0, "step": 16047 }, { "epoch": 2.0414705508205064, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.342082977294922, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8744246959686279, "num_tokens": 612265946.0, "step": 16048 }, { "epoch": 2.041597761099097, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.329977989196777, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8891092538833618, "num_tokens": 612301246.0, "step": 16049 }, { "epoch": 2.0417249713776875, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.362327575683594, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8585540652275085, "num_tokens": 612337789.0, "step": 16050 }, { "epoch": 2.041852181656278, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.386590957641602, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.8864901661872864, "num_tokens": 612372391.0, "step": 16051 }, { "epoch": 2.0419793919348685, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.320108413696289, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8775811195373535, "num_tokens": 612408806.0, "step": 16052 }, { "epoch": 2.0421066022134586, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.423775672912598, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8833005428314209, "num_tokens": 612446349.0, "step": 16053 }, { "epoch": 2.042233812492049, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.364446640014648, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8867619037628174, "num_tokens": 612489899.0, "step": 16054 }, { "epoch": 2.0423610227706397, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.31602668762207, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8675434589385986, "num_tokens": 612532314.0, "step": 16055 }, { "epoch": 2.04248823304923, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.38153076171875, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8809175491333008, "num_tokens": 612568033.0, "step": 16056 }, { "epoch": 2.0426154433278207, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.398380279541016, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8815441131591797, "num_tokens": 612605160.0, "step": 16057 }, { "epoch": 2.0427426536064113, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.31682014465332, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8621262311935425, "num_tokens": 612647851.0, "step": 16058 }, { "epoch": 2.042869863885002, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.35669231414795, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8738194108009338, "num_tokens": 612684770.0, "step": 16059 }, { "epoch": 2.0429970741635923, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.419350624084473, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8624899387359619, "num_tokens": 612721689.0, "step": 16060 }, { "epoch": 2.043124284442183, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.354758262634277, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.872833788394928, "num_tokens": 612757256.0, "step": 16061 }, { "epoch": 2.0432514947207734, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.342852592468262, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8798187971115112, "num_tokens": 612797806.0, "step": 16062 }, { "epoch": 2.043378704999364, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.343570709228516, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.884970486164093, "num_tokens": 612833507.0, "step": 16063 }, { "epoch": 2.0435059152779544, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.284215927124023, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8794910907745361, "num_tokens": 612872920.0, "step": 16064 }, { "epoch": 2.043633125556545, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.324650764465332, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8799288272857666, "num_tokens": 612909534.0, "step": 16065 }, { "epoch": 2.0437603358351355, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.340265274047852, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.870495080947876, "num_tokens": 612950531.0, "step": 16066 }, { "epoch": 2.043887546113726, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.328666687011719, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8754480481147766, "num_tokens": 612984638.0, "step": 16067 }, { "epoch": 2.0440147563923166, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.314961433410645, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8861993551254272, "num_tokens": 613020632.0, "step": 16068 }, { "epoch": 2.044141966670907, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.347661972045898, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8774944543838501, "num_tokens": 613057381.0, "step": 16069 }, { "epoch": 2.0442691769494976, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.305781364440918, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8727566003799438, "num_tokens": 613094287.0, "step": 16070 }, { "epoch": 2.044396387228088, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.440648078918457, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8802982568740845, "num_tokens": 613126315.0, "step": 16071 }, { "epoch": 2.0445235975066787, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.306424140930176, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.870073676109314, "num_tokens": 613160970.0, "step": 16072 }, { "epoch": 2.044650807785269, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.360939025878906, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8762891292572021, "num_tokens": 613204685.0, "step": 16073 }, { "epoch": 2.0447780180638597, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.378450393676758, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8831313848495483, "num_tokens": 613242394.0, "step": 16074 }, { "epoch": 2.0449052283424503, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.341538429260254, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8819451332092285, "num_tokens": 613283517.0, "step": 16075 }, { "epoch": 2.045032438621041, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.373295783996582, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8798075914382935, "num_tokens": 613317237.0, "step": 16076 }, { "epoch": 2.045159648899631, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.334810256958008, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8760027885437012, "num_tokens": 613355741.0, "step": 16077 }, { "epoch": 2.0452868591782214, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.304136276245117, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8646040558815002, "num_tokens": 613392131.0, "step": 16078 }, { "epoch": 2.045414069456812, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.334688186645508, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8647520542144775, "num_tokens": 613433687.0, "step": 16079 }, { "epoch": 2.0455412797354025, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.277592658996582, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8778764009475708, "num_tokens": 613472882.0, "step": 16080 }, { "epoch": 2.045668490013993, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.35004997253418, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8791152238845825, "num_tokens": 613515869.0, "step": 16081 }, { "epoch": 2.0457957002925835, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.371397972106934, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.875363290309906, "num_tokens": 613551835.0, "step": 16082 }, { "epoch": 2.045922910571174, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.295791625976562, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.87171471118927, "num_tokens": 613592277.0, "step": 16083 }, { "epoch": 2.0460501208497646, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.351638793945312, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8744738101959229, "num_tokens": 613625594.0, "step": 16084 }, { "epoch": 2.046177331128355, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.37147331237793, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8765658736228943, "num_tokens": 613659924.0, "step": 16085 }, { "epoch": 2.0463045414069456, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.387724876403809, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8571486473083496, "num_tokens": 613697067.0, "step": 16086 }, { "epoch": 2.046431751685536, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.360136985778809, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8808708786964417, "num_tokens": 613738211.0, "step": 16087 }, { "epoch": 2.0465589619641267, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.313395500183105, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8791344165802002, "num_tokens": 613782590.0, "step": 16088 }, { "epoch": 2.046686172242717, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.367751121520996, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8810707330703735, "num_tokens": 613825770.0, "step": 16089 }, { "epoch": 2.0468133825213077, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.323542594909668, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8799201250076294, "num_tokens": 613864025.0, "step": 16090 }, { "epoch": 2.0469405927998983, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.437823295593262, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8766639232635498, "num_tokens": 613907181.0, "step": 16091 }, { "epoch": 2.047067803078489, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.317010879516602, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8657498359680176, "num_tokens": 613951258.0, "step": 16092 }, { "epoch": 2.0471950133570793, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.326087951660156, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8718928694725037, "num_tokens": 613991221.0, "step": 16093 }, { "epoch": 2.04732222363567, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.399303436279297, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8841682076454163, "num_tokens": 614026862.0, "step": 16094 }, { "epoch": 2.0474494339142604, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.338225364685059, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8650340437889099, "num_tokens": 614067560.0, "step": 16095 }, { "epoch": 2.047576644192851, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.32766342163086, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8784275054931641, "num_tokens": 614106974.0, "step": 16096 }, { "epoch": 2.0477038544714414, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.35657787322998, "learning_rate": 1e-06, "loss": 0.312, "mean_token_accuracy": 0.9007404446601868, "num_tokens": 614145641.0, "step": 16097 }, { "epoch": 2.047831064750032, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.371698379516602, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.877828061580658, "num_tokens": 614181476.0, "step": 16098 }, { "epoch": 2.0479582750286225, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.378369331359863, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8692755699157715, "num_tokens": 614219212.0, "step": 16099 }, { "epoch": 2.048085485307213, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.394661903381348, "learning_rate": 1e-06, "loss": 0.4835, "mean_token_accuracy": 0.8456684350967407, "num_tokens": 614264017.0, "step": 16100 }, { "epoch": 2.0482126955858035, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.368518829345703, "learning_rate": 1e-06, "loss": 0.449, "mean_token_accuracy": 0.8594194650650024, "num_tokens": 614299031.0, "step": 16101 }, { "epoch": 2.0483399058643936, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.362846374511719, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8792715072631836, "num_tokens": 614334387.0, "step": 16102 }, { "epoch": 2.048467116142984, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.361859321594238, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8755877017974854, "num_tokens": 614372438.0, "step": 16103 }, { "epoch": 2.0485943264215747, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.329092025756836, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8704689741134644, "num_tokens": 614409976.0, "step": 16104 }, { "epoch": 2.048721536700165, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.338075637817383, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8828654289245605, "num_tokens": 614448740.0, "step": 16105 }, { "epoch": 2.0488487469787557, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.32361125946045, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8776929378509521, "num_tokens": 614484644.0, "step": 16106 }, { "epoch": 2.0489759572573463, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.376205444335938, "learning_rate": 1e-06, "loss": 0.4528, "mean_token_accuracy": 0.85491943359375, "num_tokens": 614521210.0, "step": 16107 }, { "epoch": 2.049103167535937, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.288469314575195, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8691489696502686, "num_tokens": 614561783.0, "step": 16108 }, { "epoch": 2.0492303778145273, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.419395446777344, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8749074935913086, "num_tokens": 614598643.0, "step": 16109 }, { "epoch": 2.049357588093118, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.346826553344727, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8742286562919617, "num_tokens": 614633520.0, "step": 16110 }, { "epoch": 2.0494847983717084, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.320168495178223, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.860822319984436, "num_tokens": 614673266.0, "step": 16111 }, { "epoch": 2.049612008650299, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.420211791992188, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8649955987930298, "num_tokens": 614707201.0, "step": 16112 }, { "epoch": 2.0497392189288894, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.330822944641113, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8790962100028992, "num_tokens": 614746542.0, "step": 16113 }, { "epoch": 2.04986642920748, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.38961410522461, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8633118867874146, "num_tokens": 614781804.0, "step": 16114 }, { "epoch": 2.0499936394860705, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.35987663269043, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8736697435379028, "num_tokens": 614817281.0, "step": 16115 }, { "epoch": 2.050120849764661, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.370508193969727, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8640600442886353, "num_tokens": 614853666.0, "step": 16116 }, { "epoch": 2.0502480600432516, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.300860404968262, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8750089406967163, "num_tokens": 614891221.0, "step": 16117 }, { "epoch": 2.050375270321842, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.40194320678711, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8813201189041138, "num_tokens": 614925956.0, "step": 16118 }, { "epoch": 2.0505024806004326, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.425020217895508, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8797260522842407, "num_tokens": 614959934.0, "step": 16119 }, { "epoch": 2.050629690879023, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.331066131591797, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8704915642738342, "num_tokens": 615000579.0, "step": 16120 }, { "epoch": 2.0507569011576137, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.43442440032959, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8670558929443359, "num_tokens": 615037371.0, "step": 16121 }, { "epoch": 2.050884111436204, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.360786437988281, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.863916277885437, "num_tokens": 615075948.0, "step": 16122 }, { "epoch": 2.0510113217147947, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.421247482299805, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8863611221313477, "num_tokens": 615114619.0, "step": 16123 }, { "epoch": 2.0511385319933853, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.38197135925293, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8745837211608887, "num_tokens": 615149260.0, "step": 16124 }, { "epoch": 2.051265742271976, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.43013858795166, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8777065277099609, "num_tokens": 615183284.0, "step": 16125 }, { "epoch": 2.0513929525505663, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.406240463256836, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8798080682754517, "num_tokens": 615220015.0, "step": 16126 }, { "epoch": 2.0515201628291564, "ewc_loss": 0.0341796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.381718635559082, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8688867092132568, "num_tokens": 615255614.0, "step": 16127 }, { "epoch": 2.051647373107747, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.427079200744629, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8751682639122009, "num_tokens": 615290232.0, "step": 16128 }, { "epoch": 2.0517745833863374, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.369355201721191, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.87917160987854, "num_tokens": 615328204.0, "step": 16129 }, { "epoch": 2.051901793664928, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.47802734375, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8645514845848083, "num_tokens": 615362846.0, "step": 16130 }, { "epoch": 2.0520290039435185, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.371113777160645, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8732935786247253, "num_tokens": 615405404.0, "step": 16131 }, { "epoch": 2.052156214222109, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.383612632751465, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8632744550704956, "num_tokens": 615439835.0, "step": 16132 }, { "epoch": 2.0522834245006996, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.347760200500488, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8842926025390625, "num_tokens": 615477832.0, "step": 16133 }, { "epoch": 2.05241063477929, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.387593269348145, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8663480877876282, "num_tokens": 615518089.0, "step": 16134 }, { "epoch": 2.0525378450578806, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.366044044494629, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8791441917419434, "num_tokens": 615557286.0, "step": 16135 }, { "epoch": 2.052665055336471, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.42599105834961, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8775973320007324, "num_tokens": 615593284.0, "step": 16136 }, { "epoch": 2.0527922656150617, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.368094444274902, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8692993521690369, "num_tokens": 615633579.0, "step": 16137 }, { "epoch": 2.052919475893652, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.367622375488281, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8780511021614075, "num_tokens": 615680135.0, "step": 16138 }, { "epoch": 2.0530466861722427, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.387975692749023, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8629738688468933, "num_tokens": 615714801.0, "step": 16139 }, { "epoch": 2.0531738964508333, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.353150367736816, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8657231330871582, "num_tokens": 615750538.0, "step": 16140 }, { "epoch": 2.053301106729424, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.410354614257812, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8796923160552979, "num_tokens": 615789465.0, "step": 16141 }, { "epoch": 2.0534283170080143, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.33846664428711, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8599618673324585, "num_tokens": 615832553.0, "step": 16142 }, { "epoch": 2.053555527286605, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.412631034851074, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.869664192199707, "num_tokens": 615868341.0, "step": 16143 }, { "epoch": 2.0536827375651954, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.359959602355957, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8829877972602844, "num_tokens": 615904324.0, "step": 16144 }, { "epoch": 2.053809947843786, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.348519325256348, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8803638219833374, "num_tokens": 615947656.0, "step": 16145 }, { "epoch": 2.0539371581223764, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.403302192687988, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8684574961662292, "num_tokens": 615989050.0, "step": 16146 }, { "epoch": 2.054064368400967, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.412212371826172, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8776992559432983, "num_tokens": 616027638.0, "step": 16147 }, { "epoch": 2.0541915786795575, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.398646354675293, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8816958069801331, "num_tokens": 616066392.0, "step": 16148 }, { "epoch": 2.054318788958148, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.388644218444824, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8830996155738831, "num_tokens": 616108303.0, "step": 16149 }, { "epoch": 2.0544459992367385, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.373875617980957, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8661605715751648, "num_tokens": 616154174.0, "step": 16150 }, { "epoch": 2.0545732095153286, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.357148170471191, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.871627688407898, "num_tokens": 616193954.0, "step": 16151 }, { "epoch": 2.054700419793919, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.400331497192383, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8787009119987488, "num_tokens": 616232771.0, "step": 16152 }, { "epoch": 2.0548276300725097, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.382950782775879, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8843809962272644, "num_tokens": 616265151.0, "step": 16153 }, { "epoch": 2.0549548403511, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.38008975982666, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8683789968490601, "num_tokens": 616303886.0, "step": 16154 }, { "epoch": 2.0550820506296907, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.353157997131348, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8830289840698242, "num_tokens": 616340040.0, "step": 16155 }, { "epoch": 2.0552092609082813, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.390912055969238, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8789923191070557, "num_tokens": 616371329.0, "step": 16156 }, { "epoch": 2.055336471186872, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.397497177124023, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8881406784057617, "num_tokens": 616406383.0, "step": 16157 }, { "epoch": 2.0554636814654623, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.344026565551758, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8613237738609314, "num_tokens": 616445600.0, "step": 16158 }, { "epoch": 2.055590891744053, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 13.323223114013672, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8665053248405457, "num_tokens": 616485376.0, "step": 16159 }, { "epoch": 2.0557181020226434, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.39299201965332, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8654824495315552, "num_tokens": 616527104.0, "step": 16160 }, { "epoch": 2.055845312301234, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.36622142791748, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8763958215713501, "num_tokens": 616567557.0, "step": 16161 }, { "epoch": 2.0559725225798244, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.375157356262207, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8795496225357056, "num_tokens": 616601612.0, "step": 16162 }, { "epoch": 2.056099732858415, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.371871948242188, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.871002733707428, "num_tokens": 616638019.0, "step": 16163 }, { "epoch": 2.0562269431370055, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.331372261047363, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8712693452835083, "num_tokens": 616675224.0, "step": 16164 }, { "epoch": 2.056354153415596, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.429807662963867, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8723203539848328, "num_tokens": 616713628.0, "step": 16165 }, { "epoch": 2.0564813636941865, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.345656394958496, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8716540336608887, "num_tokens": 616755985.0, "step": 16166 }, { "epoch": 2.056608573972777, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.43887710571289, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8742315173149109, "num_tokens": 616786564.0, "step": 16167 }, { "epoch": 2.0567357842513676, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.35983657836914, "learning_rate": 1e-06, "loss": 0.4823, "mean_token_accuracy": 0.846692681312561, "num_tokens": 616823285.0, "step": 16168 }, { "epoch": 2.056862994529958, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.384289741516113, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8812595009803772, "num_tokens": 616864465.0, "step": 16169 }, { "epoch": 2.0569902048085487, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.370150566101074, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8647938966751099, "num_tokens": 616906670.0, "step": 16170 }, { "epoch": 2.057117415087139, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.33526611328125, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8724374771118164, "num_tokens": 616947656.0, "step": 16171 }, { "epoch": 2.0572446253657297, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.458050727844238, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8666645288467407, "num_tokens": 616980161.0, "step": 16172 }, { "epoch": 2.0573718356443202, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.351019859313965, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8728088140487671, "num_tokens": 617015281.0, "step": 16173 }, { "epoch": 2.0574990459229108, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.391199111938477, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8749948740005493, "num_tokens": 617054174.0, "step": 16174 }, { "epoch": 2.057626256201501, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.352131843566895, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8726189732551575, "num_tokens": 617095160.0, "step": 16175 }, { "epoch": 2.0577534664800914, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.379425048828125, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8781591057777405, "num_tokens": 617130781.0, "step": 16176 }, { "epoch": 2.057880676758682, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.421553611755371, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.871379017829895, "num_tokens": 617169473.0, "step": 16177 }, { "epoch": 2.0580078870372724, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.377140998840332, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8675154447555542, "num_tokens": 617212208.0, "step": 16178 }, { "epoch": 2.058135097315863, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.34009838104248, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8763241171836853, "num_tokens": 617246055.0, "step": 16179 }, { "epoch": 2.0582623075944535, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.388081550598145, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.869165301322937, "num_tokens": 617285614.0, "step": 16180 }, { "epoch": 2.058389517873044, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.372936248779297, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8823808431625366, "num_tokens": 617323401.0, "step": 16181 }, { "epoch": 2.0585167281516346, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.382792472839355, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8602322340011597, "num_tokens": 617372550.0, "step": 16182 }, { "epoch": 2.058643938430225, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.388405799865723, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.880087673664093, "num_tokens": 617407894.0, "step": 16183 }, { "epoch": 2.0587711487088156, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.397333145141602, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8658562302589417, "num_tokens": 617450117.0, "step": 16184 }, { "epoch": 2.058898358987406, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.407514572143555, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8788371086120605, "num_tokens": 617485885.0, "step": 16185 }, { "epoch": 2.0590255692659967, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.380282402038574, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8574652671813965, "num_tokens": 617520511.0, "step": 16186 }, { "epoch": 2.059152779544587, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39803695678711, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8673055171966553, "num_tokens": 617558288.0, "step": 16187 }, { "epoch": 2.0592799898231777, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.339227676391602, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8723281025886536, "num_tokens": 617602991.0, "step": 16188 }, { "epoch": 2.0594072001017683, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39576244354248, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8706718683242798, "num_tokens": 617639629.0, "step": 16189 }, { "epoch": 2.059534410380359, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.37674617767334, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8666399121284485, "num_tokens": 617676218.0, "step": 16190 }, { "epoch": 2.0596616206589493, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.339920043945312, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8664314150810242, "num_tokens": 617716270.0, "step": 16191 }, { "epoch": 2.05978883093754, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39742374420166, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8827142119407654, "num_tokens": 617755559.0, "step": 16192 }, { "epoch": 2.0599160412161304, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3618745803833, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.874607503414154, "num_tokens": 617793522.0, "step": 16193 }, { "epoch": 2.060043251494721, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.393938064575195, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8774948120117188, "num_tokens": 617828005.0, "step": 16194 }, { "epoch": 2.0601704617733114, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.354286193847656, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8695798516273499, "num_tokens": 617863824.0, "step": 16195 }, { "epoch": 2.060297672051902, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.394508361816406, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.869153618812561, "num_tokens": 617903648.0, "step": 16196 }, { "epoch": 2.0604248823304925, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.306360244750977, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8705947995185852, "num_tokens": 617946229.0, "step": 16197 }, { "epoch": 2.060552092609083, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.34499740600586, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8735226392745972, "num_tokens": 617990486.0, "step": 16198 }, { "epoch": 2.0606793028876735, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.321318626403809, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8848717212677002, "num_tokens": 618031743.0, "step": 16199 }, { "epoch": 2.0608065131662636, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.385616302490234, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8706341981887817, "num_tokens": 618070181.0, "step": 16200 }, { "epoch": 2.060933723444854, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.405320167541504, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8801374435424805, "num_tokens": 618106396.0, "step": 16201 }, { "epoch": 2.0610609337234447, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.327926635742188, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8780367970466614, "num_tokens": 618148281.0, "step": 16202 }, { "epoch": 2.061188144002035, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.448959350585938, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8668553829193115, "num_tokens": 618182721.0, "step": 16203 }, { "epoch": 2.0613153542806257, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.326998710632324, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8882433772087097, "num_tokens": 618219497.0, "step": 16204 }, { "epoch": 2.0614425645592163, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.386512756347656, "learning_rate": 1e-06, "loss": 0.3525, "mean_token_accuracy": 0.8899694681167603, "num_tokens": 618258643.0, "step": 16205 }, { "epoch": 2.061569774837807, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.390405654907227, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8880958557128906, "num_tokens": 618292682.0, "step": 16206 }, { "epoch": 2.0616969851163973, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.328376770019531, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8585941791534424, "num_tokens": 618331375.0, "step": 16207 }, { "epoch": 2.061824195394988, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.383645057678223, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8785744905471802, "num_tokens": 618367452.0, "step": 16208 }, { "epoch": 2.0619514056735784, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.381874084472656, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8718915581703186, "num_tokens": 618410470.0, "step": 16209 }, { "epoch": 2.062078615952169, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.344558715820312, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.865634560585022, "num_tokens": 618447524.0, "step": 16210 }, { "epoch": 2.0622058262307594, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.421622276306152, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8541834950447083, "num_tokens": 618481793.0, "step": 16211 }, { "epoch": 2.06233303650935, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.403996467590332, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8806995153427124, "num_tokens": 618517097.0, "step": 16212 }, { "epoch": 2.0624602467879405, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.376334190368652, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8716965913772583, "num_tokens": 618554221.0, "step": 16213 }, { "epoch": 2.062587457066531, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.322161674499512, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8834151029586792, "num_tokens": 618591279.0, "step": 16214 }, { "epoch": 2.0627146673451215, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.38546371459961, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.860985279083252, "num_tokens": 618628168.0, "step": 16215 }, { "epoch": 2.062841877623712, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.368210792541504, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.878548264503479, "num_tokens": 618664580.0, "step": 16216 }, { "epoch": 2.0629690879023026, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.434178352355957, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8626474142074585, "num_tokens": 618699007.0, "step": 16217 }, { "epoch": 2.063096298180893, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.353754997253418, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8747327327728271, "num_tokens": 618736630.0, "step": 16218 }, { "epoch": 2.0632235084594837, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.327472686767578, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8814629316329956, "num_tokens": 618777868.0, "step": 16219 }, { "epoch": 2.063350718738074, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.390576362609863, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8779922127723694, "num_tokens": 618819134.0, "step": 16220 }, { "epoch": 2.0634779290166647, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.336524963378906, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8736960887908936, "num_tokens": 618862231.0, "step": 16221 }, { "epoch": 2.0636051392952552, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.399530410766602, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8661469221115112, "num_tokens": 618906816.0, "step": 16222 }, { "epoch": 2.0637323495738458, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.418510437011719, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8745250105857849, "num_tokens": 618943239.0, "step": 16223 }, { "epoch": 2.0638595598524363, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.368321418762207, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.8867387771606445, "num_tokens": 618978641.0, "step": 16224 }, { "epoch": 2.0639867701310264, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.399261474609375, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.874492347240448, "num_tokens": 619020522.0, "step": 16225 }, { "epoch": 2.064113980409617, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.409753799438477, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8853344917297363, "num_tokens": 619060769.0, "step": 16226 }, { "epoch": 2.0642411906882074, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.390229225158691, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8846644759178162, "num_tokens": 619098711.0, "step": 16227 }, { "epoch": 2.064368400966798, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.400477409362793, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8697611093521118, "num_tokens": 619133673.0, "step": 16228 }, { "epoch": 2.0644956112453885, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39828872680664, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.867605984210968, "num_tokens": 619174554.0, "step": 16229 }, { "epoch": 2.064622821523979, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.397096633911133, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8631173968315125, "num_tokens": 619211095.0, "step": 16230 }, { "epoch": 2.0647500318025696, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.336480140686035, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8845869302749634, "num_tokens": 619246346.0, "step": 16231 }, { "epoch": 2.06487724208116, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.400336265563965, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8706163763999939, "num_tokens": 619280629.0, "step": 16232 }, { "epoch": 2.0650044523597506, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.387886047363281, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8540043830871582, "num_tokens": 619320928.0, "step": 16233 }, { "epoch": 2.065131662638341, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.388481140136719, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8745847344398499, "num_tokens": 619369747.0, "step": 16234 }, { "epoch": 2.0652588729169317, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.398504257202148, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8757519125938416, "num_tokens": 619408694.0, "step": 16235 }, { "epoch": 2.065386083195522, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.441673278808594, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8810825943946838, "num_tokens": 619446858.0, "step": 16236 }, { "epoch": 2.0655132934741127, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.445693969726562, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8734578490257263, "num_tokens": 619482281.0, "step": 16237 }, { "epoch": 2.0656405037527032, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.386018753051758, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8832576870918274, "num_tokens": 619520309.0, "step": 16238 }, { "epoch": 2.065767714031294, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.438870429992676, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.87773197889328, "num_tokens": 619554662.0, "step": 16239 }, { "epoch": 2.0658949243098843, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.429449081420898, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8711690902709961, "num_tokens": 619599632.0, "step": 16240 }, { "epoch": 2.066022134588475, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.417399406433105, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8863039016723633, "num_tokens": 619639414.0, "step": 16241 }, { "epoch": 2.0661493448670654, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.446859359741211, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8763437271118164, "num_tokens": 619675878.0, "step": 16242 }, { "epoch": 2.066276555145656, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.469420433044434, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8769732117652893, "num_tokens": 619709486.0, "step": 16243 }, { "epoch": 2.0664037654242464, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.388077735900879, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8746331930160522, "num_tokens": 619753576.0, "step": 16244 }, { "epoch": 2.066530975702837, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.380460739135742, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8750385046005249, "num_tokens": 619792520.0, "step": 16245 }, { "epoch": 2.0666581859814275, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.496603012084961, "learning_rate": 1e-06, "loss": 0.4824, "mean_token_accuracy": 0.8536272644996643, "num_tokens": 619827871.0, "step": 16246 }, { "epoch": 2.066785396260018, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.348923683166504, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.86920565366745, "num_tokens": 619864527.0, "step": 16247 }, { "epoch": 2.0669126065386085, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.437347412109375, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8755032420158386, "num_tokens": 619904934.0, "step": 16248 }, { "epoch": 2.0670398168171986, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.395317077636719, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8814881443977356, "num_tokens": 619945472.0, "step": 16249 }, { "epoch": 2.067167027095789, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.389474868774414, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8857583999633789, "num_tokens": 619981738.0, "step": 16250 }, { "epoch": 2.0672942373743797, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.417501449584961, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8758032321929932, "num_tokens": 620024593.0, "step": 16251 }, { "epoch": 2.06742144765297, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.374468803405762, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8807258009910583, "num_tokens": 620064059.0, "step": 16252 }, { "epoch": 2.0675486579315607, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.376781463623047, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8755772113800049, "num_tokens": 620100028.0, "step": 16253 }, { "epoch": 2.0676758682101513, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.457850456237793, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8872177004814148, "num_tokens": 620135181.0, "step": 16254 }, { "epoch": 2.067803078488742, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.375252723693848, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8760374784469604, "num_tokens": 620174887.0, "step": 16255 }, { "epoch": 2.0679302887673323, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.46322250366211, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.874500572681427, "num_tokens": 620212924.0, "step": 16256 }, { "epoch": 2.068057499045923, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.38225269317627, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8657357692718506, "num_tokens": 620251088.0, "step": 16257 }, { "epoch": 2.0681847093245134, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.424215316772461, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.873911440372467, "num_tokens": 620287888.0, "step": 16258 }, { "epoch": 2.068311919603104, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.417638778686523, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8727877140045166, "num_tokens": 620324937.0, "step": 16259 }, { "epoch": 2.0684391298816944, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.43604850769043, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8750970363616943, "num_tokens": 620363390.0, "step": 16260 }, { "epoch": 2.068566340160285, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442381858825684, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8817751407623291, "num_tokens": 620403707.0, "step": 16261 }, { "epoch": 2.0686935504388755, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.431119918823242, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.878288745880127, "num_tokens": 620445994.0, "step": 16262 }, { "epoch": 2.068820760717466, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.41197681427002, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8815635442733765, "num_tokens": 620480883.0, "step": 16263 }, { "epoch": 2.0689479709960565, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.411142349243164, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8866300582885742, "num_tokens": 620518935.0, "step": 16264 }, { "epoch": 2.069075181274647, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.45910358428955, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8683846592903137, "num_tokens": 620551985.0, "step": 16265 }, { "epoch": 2.0692023915532376, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.420113563537598, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8698571920394897, "num_tokens": 620592809.0, "step": 16266 }, { "epoch": 2.069329601831828, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.368328094482422, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8603013753890991, "num_tokens": 620627951.0, "step": 16267 }, { "epoch": 2.0694568121104187, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.47011947631836, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8795760869979858, "num_tokens": 620660892.0, "step": 16268 }, { "epoch": 2.069584022389009, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.335489273071289, "learning_rate": 1e-06, "loss": 0.4673, "mean_token_accuracy": 0.8514686822891235, "num_tokens": 620698165.0, "step": 16269 }, { "epoch": 2.0697112326675997, "ewc_loss": 0.034423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.46323299407959, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8652356863021851, "num_tokens": 620734758.0, "step": 16270 }, { "epoch": 2.0698384429461902, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.342880249023438, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8558067083358765, "num_tokens": 620770224.0, "step": 16271 }, { "epoch": 2.0699656532247808, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.363924980163574, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8821837902069092, "num_tokens": 620805724.0, "step": 16272 }, { "epoch": 2.070092863503371, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.385943412780762, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.863259494304657, "num_tokens": 620847622.0, "step": 16273 }, { "epoch": 2.0702200737819614, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.393899917602539, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8713567852973938, "num_tokens": 620879042.0, "step": 16274 }, { "epoch": 2.070347284060552, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.376391410827637, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8702552914619446, "num_tokens": 620922685.0, "step": 16275 }, { "epoch": 2.0704744943391424, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.368038177490234, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8687958121299744, "num_tokens": 620954854.0, "step": 16276 }, { "epoch": 2.070601704617733, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.397320747375488, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8714174032211304, "num_tokens": 620994581.0, "step": 16277 }, { "epoch": 2.0707289148963235, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.461443901062012, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8746248483657837, "num_tokens": 621030052.0, "step": 16278 }, { "epoch": 2.070856125174914, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.413314819335938, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8830727338790894, "num_tokens": 621072474.0, "step": 16279 }, { "epoch": 2.0709833354535045, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.443793296813965, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8770769834518433, "num_tokens": 621104365.0, "step": 16280 }, { "epoch": 2.071110545732095, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.468954086303711, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8788148164749146, "num_tokens": 621140497.0, "step": 16281 }, { "epoch": 2.0712377560106856, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.380990028381348, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8765436410903931, "num_tokens": 621177431.0, "step": 16282 }, { "epoch": 2.071364966289276, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.440916061401367, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8767135143280029, "num_tokens": 621212874.0, "step": 16283 }, { "epoch": 2.0714921765678667, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.373987197875977, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8791648745536804, "num_tokens": 621249927.0, "step": 16284 }, { "epoch": 2.071619386846457, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.469143867492676, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8496493101119995, "num_tokens": 621280794.0, "step": 16285 }, { "epoch": 2.0717465971250477, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.42992877960205, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8576328754425049, "num_tokens": 621321167.0, "step": 16286 }, { "epoch": 2.0718738074036382, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.409939765930176, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8847396969795227, "num_tokens": 621357962.0, "step": 16287 }, { "epoch": 2.0720010176822288, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.395072937011719, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.8920495510101318, "num_tokens": 621398933.0, "step": 16288 }, { "epoch": 2.0721282279608193, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.385435104370117, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8926817178726196, "num_tokens": 621442334.0, "step": 16289 }, { "epoch": 2.07225543823941, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.364642143249512, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8763085603713989, "num_tokens": 621478706.0, "step": 16290 }, { "epoch": 2.0723826485180004, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.437088012695312, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8793658018112183, "num_tokens": 621513911.0, "step": 16291 }, { "epoch": 2.072509858796591, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.441679954528809, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8689776659011841, "num_tokens": 621554080.0, "step": 16292 }, { "epoch": 2.0726370690751814, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39206600189209, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8766392469406128, "num_tokens": 621592059.0, "step": 16293 }, { "epoch": 2.072764279353772, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.427430152893066, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8855891227722168, "num_tokens": 621629626.0, "step": 16294 }, { "epoch": 2.0728914896323625, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.405889511108398, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8784434795379639, "num_tokens": 621671341.0, "step": 16295 }, { "epoch": 2.073018699910953, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.396105766296387, "learning_rate": 1e-06, "loss": 0.3417, "mean_token_accuracy": 0.8884977698326111, "num_tokens": 621708756.0, "step": 16296 }, { "epoch": 2.0731459101895435, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.411787986755371, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8790898323059082, "num_tokens": 621748009.0, "step": 16297 }, { "epoch": 2.0732731204681336, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.38133716583252, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8828682899475098, "num_tokens": 621788695.0, "step": 16298 }, { "epoch": 2.073400330746724, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.370193481445312, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8732675313949585, "num_tokens": 621828099.0, "step": 16299 }, { "epoch": 2.0735275410253147, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.430129051208496, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8571649789810181, "num_tokens": 621867033.0, "step": 16300 }, { "epoch": 2.073654751303905, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.350163459777832, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8638785481452942, "num_tokens": 621907435.0, "step": 16301 }, { "epoch": 2.0737819615824957, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.357177734375, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.865045428276062, "num_tokens": 621948995.0, "step": 16302 }, { "epoch": 2.0739091718610863, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.359631538391113, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8855310082435608, "num_tokens": 621991109.0, "step": 16303 }, { "epoch": 2.074036382139677, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.365534782409668, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8807694315910339, "num_tokens": 622026585.0, "step": 16304 }, { "epoch": 2.0741635924182673, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.426803588867188, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8837374448776245, "num_tokens": 622065208.0, "step": 16305 }, { "epoch": 2.074290802696858, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.381117820739746, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.865570068359375, "num_tokens": 622104838.0, "step": 16306 }, { "epoch": 2.0744180129754484, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.416691780090332, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8653991222381592, "num_tokens": 622136879.0, "step": 16307 }, { "epoch": 2.074545223254039, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.378946304321289, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8852112293243408, "num_tokens": 622175938.0, "step": 16308 }, { "epoch": 2.0746724335326294, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.359782218933105, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8830685615539551, "num_tokens": 622216663.0, "step": 16309 }, { "epoch": 2.07479964381122, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.455525398254395, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8832608461380005, "num_tokens": 622250450.0, "step": 16310 }, { "epoch": 2.0749268540898105, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.409299850463867, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8670021891593933, "num_tokens": 622291919.0, "step": 16311 }, { "epoch": 2.075054064368401, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.362167358398438, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8737388849258423, "num_tokens": 622325892.0, "step": 16312 }, { "epoch": 2.0751812746469915, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.435826301574707, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8743268847465515, "num_tokens": 622356966.0, "step": 16313 }, { "epoch": 2.075308484925582, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.401026725769043, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8779274225234985, "num_tokens": 622389931.0, "step": 16314 }, { "epoch": 2.0754356952041726, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.38804817199707, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8734108209609985, "num_tokens": 622427817.0, "step": 16315 }, { "epoch": 2.075562905482763, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.365580558776855, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8654450178146362, "num_tokens": 622470353.0, "step": 16316 }, { "epoch": 2.0756901157613536, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.367148399353027, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.89002525806427, "num_tokens": 622507577.0, "step": 16317 }, { "epoch": 2.075817326039944, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.382226943969727, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8703368902206421, "num_tokens": 622544746.0, "step": 16318 }, { "epoch": 2.0759445363185347, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.406160354614258, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8680062294006348, "num_tokens": 622583688.0, "step": 16319 }, { "epoch": 2.0760717465971252, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.375558853149414, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8746140003204346, "num_tokens": 622619911.0, "step": 16320 }, { "epoch": 2.0761989568757158, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.38443374633789, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8799008131027222, "num_tokens": 622665732.0, "step": 16321 }, { "epoch": 2.0763261671543063, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.339033126831055, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8816177845001221, "num_tokens": 622705746.0, "step": 16322 }, { "epoch": 2.0764533774328964, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.450756072998047, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8757601380348206, "num_tokens": 622745481.0, "step": 16323 }, { "epoch": 2.076580587711487, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.380776405334473, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8771543502807617, "num_tokens": 622786910.0, "step": 16324 }, { "epoch": 2.0767077979900774, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.406868934631348, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8705446124076843, "num_tokens": 622826976.0, "step": 16325 }, { "epoch": 2.076835008268668, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442878723144531, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8755086064338684, "num_tokens": 622858945.0, "step": 16326 }, { "epoch": 2.0769622185472585, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.392143249511719, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8862186670303345, "num_tokens": 622893229.0, "step": 16327 }, { "epoch": 2.077089428825849, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.388654708862305, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8846800923347473, "num_tokens": 622933701.0, "step": 16328 }, { "epoch": 2.0772166391044395, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.425373077392578, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8605964183807373, "num_tokens": 622971226.0, "step": 16329 }, { "epoch": 2.07734384938303, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.367908477783203, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8708208799362183, "num_tokens": 623009232.0, "step": 16330 }, { "epoch": 2.0774710596616206, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.376336097717285, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8696428537368774, "num_tokens": 623048005.0, "step": 16331 }, { "epoch": 2.077598269940211, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.373602867126465, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8741410970687866, "num_tokens": 623088474.0, "step": 16332 }, { "epoch": 2.0777254802188017, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.335186004638672, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8758796453475952, "num_tokens": 623123700.0, "step": 16333 }, { "epoch": 2.077852690497392, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.37999439239502, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8776944875717163, "num_tokens": 623161406.0, "step": 16334 }, { "epoch": 2.0779799007759827, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.403717994689941, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8646159172058105, "num_tokens": 623199770.0, "step": 16335 }, { "epoch": 2.0781071110545732, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.421966552734375, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8772022724151611, "num_tokens": 623234635.0, "step": 16336 }, { "epoch": 2.0782343213331638, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.36438274383545, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8765413761138916, "num_tokens": 623273695.0, "step": 16337 }, { "epoch": 2.0783615316117543, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.41207218170166, "learning_rate": 1e-06, "loss": 0.4434, "mean_token_accuracy": 0.8599761724472046, "num_tokens": 623315278.0, "step": 16338 }, { "epoch": 2.078488741890345, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.417084693908691, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8834333419799805, "num_tokens": 623351753.0, "step": 16339 }, { "epoch": 2.0786159521689354, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.414949417114258, "learning_rate": 1e-06, "loss": 0.4657, "mean_token_accuracy": 0.8568086624145508, "num_tokens": 623387812.0, "step": 16340 }, { "epoch": 2.078743162447526, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.438714981079102, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8786618709564209, "num_tokens": 623426784.0, "step": 16341 }, { "epoch": 2.0788703727261164, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.398828506469727, "learning_rate": 1e-06, "loss": 0.479, "mean_token_accuracy": 0.8523870706558228, "num_tokens": 623463362.0, "step": 16342 }, { "epoch": 2.078997583004707, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.415505409240723, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8678053617477417, "num_tokens": 623500108.0, "step": 16343 }, { "epoch": 2.0791247932832975, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.32171630859375, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8708989024162292, "num_tokens": 623537159.0, "step": 16344 }, { "epoch": 2.079252003561888, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.348926544189453, "learning_rate": 1e-06, "loss": 0.3397, "mean_token_accuracy": 0.8932014107704163, "num_tokens": 623576256.0, "step": 16345 }, { "epoch": 2.0793792138404785, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.444971084594727, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8714029788970947, "num_tokens": 623612454.0, "step": 16346 }, { "epoch": 2.0795064241190686, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.411808013916016, "learning_rate": 1e-06, "loss": 0.3485, "mean_token_accuracy": 0.8872243165969849, "num_tokens": 623654392.0, "step": 16347 }, { "epoch": 2.079633634397659, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.446745872497559, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8565011620521545, "num_tokens": 623697350.0, "step": 16348 }, { "epoch": 2.0797608446762497, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.407659530639648, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8784127831459045, "num_tokens": 623736272.0, "step": 16349 }, { "epoch": 2.07988805495484, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.415404319763184, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8647960424423218, "num_tokens": 623781319.0, "step": 16350 }, { "epoch": 2.0800152652334307, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.38635540008545, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.875527024269104, "num_tokens": 623821944.0, "step": 16351 }, { "epoch": 2.0801424755120212, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.433996200561523, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8697468042373657, "num_tokens": 623858399.0, "step": 16352 }, { "epoch": 2.0802696857906118, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.383469581604004, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8761826753616333, "num_tokens": 623897839.0, "step": 16353 }, { "epoch": 2.0803968960692023, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.429898262023926, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.859340488910675, "num_tokens": 623933127.0, "step": 16354 }, { "epoch": 2.080524106347793, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.450746536254883, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8740994930267334, "num_tokens": 623972140.0, "step": 16355 }, { "epoch": 2.0806513166263834, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.404996871948242, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.86363685131073, "num_tokens": 624008875.0, "step": 16356 }, { "epoch": 2.080778526904974, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.392919540405273, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8708513975143433, "num_tokens": 624045162.0, "step": 16357 }, { "epoch": 2.0809057371835644, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.3735990524292, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8837047815322876, "num_tokens": 624085275.0, "step": 16358 }, { "epoch": 2.081032947462155, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.391486167907715, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8720640540122986, "num_tokens": 624127479.0, "step": 16359 }, { "epoch": 2.0811601577407455, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.395657539367676, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8766462802886963, "num_tokens": 624165231.0, "step": 16360 }, { "epoch": 2.081287368019336, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.382543563842773, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8628596067428589, "num_tokens": 624205109.0, "step": 16361 }, { "epoch": 2.0814145782979265, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.44063949584961, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8690823912620544, "num_tokens": 624241849.0, "step": 16362 }, { "epoch": 2.081541788576517, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.376983642578125, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8754699230194092, "num_tokens": 624274174.0, "step": 16363 }, { "epoch": 2.0816689988551076, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.398165702819824, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8716931343078613, "num_tokens": 624311526.0, "step": 16364 }, { "epoch": 2.081796209133698, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.415822982788086, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8610703945159912, "num_tokens": 624346657.0, "step": 16365 }, { "epoch": 2.0819234194122886, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.383066177368164, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8578450679779053, "num_tokens": 624386563.0, "step": 16366 }, { "epoch": 2.082050629690879, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.391624450683594, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8840253949165344, "num_tokens": 624427299.0, "step": 16367 }, { "epoch": 2.0821778399694697, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.421762466430664, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8567091226577759, "num_tokens": 624461253.0, "step": 16368 }, { "epoch": 2.0823050502480602, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.395044326782227, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8831594586372375, "num_tokens": 624498699.0, "step": 16369 }, { "epoch": 2.0824322605266508, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.358403205871582, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.871857762336731, "num_tokens": 624534655.0, "step": 16370 }, { "epoch": 2.082559470805241, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.34736156463623, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8771421909332275, "num_tokens": 624575342.0, "step": 16371 }, { "epoch": 2.0826866810838314, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.381190299987793, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8788938522338867, "num_tokens": 624608493.0, "step": 16372 }, { "epoch": 2.082813891362422, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.367122650146484, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8560887575149536, "num_tokens": 624655777.0, "step": 16373 }, { "epoch": 2.0829411016410124, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.303986549377441, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8841355443000793, "num_tokens": 624699349.0, "step": 16374 }, { "epoch": 2.083068311919603, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.449119567871094, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8818379044532776, "num_tokens": 624730146.0, "step": 16375 }, { "epoch": 2.0831955221981935, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.334087371826172, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8658266663551331, "num_tokens": 624773186.0, "step": 16376 }, { "epoch": 2.083322732476784, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.362284660339355, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8720653653144836, "num_tokens": 624810256.0, "step": 16377 }, { "epoch": 2.0834499427553745, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.393184661865234, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8640604019165039, "num_tokens": 624848166.0, "step": 16378 }, { "epoch": 2.083577153033965, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.334209442138672, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8727070689201355, "num_tokens": 624887601.0, "step": 16379 }, { "epoch": 2.0837043633125556, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.456738471984863, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8738632202148438, "num_tokens": 624921160.0, "step": 16380 }, { "epoch": 2.083831573591146, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.347726821899414, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8670433163642883, "num_tokens": 624959053.0, "step": 16381 }, { "epoch": 2.0839587838697367, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.432307243347168, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8867241144180298, "num_tokens": 624995272.0, "step": 16382 }, { "epoch": 2.084085994148327, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.382132530212402, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8892885446548462, "num_tokens": 625033388.0, "step": 16383 }, { "epoch": 2.0842132044269177, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.4240083694458, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8626564741134644, "num_tokens": 625071103.0, "step": 16384 }, { "epoch": 2.0843404147055082, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.394804000854492, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8855628967285156, "num_tokens": 625109562.0, "step": 16385 }, { "epoch": 2.0844676249840988, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.409249305725098, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8794408440589905, "num_tokens": 625147438.0, "step": 16386 }, { "epoch": 2.0845948352626893, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.40211296081543, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8827475905418396, "num_tokens": 625184293.0, "step": 16387 }, { "epoch": 2.08472204554128, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.419646263122559, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8757244944572449, "num_tokens": 625218201.0, "step": 16388 }, { "epoch": 2.0848492558198704, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.417073249816895, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8718373775482178, "num_tokens": 625252072.0, "step": 16389 }, { "epoch": 2.084976466098461, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.438719749450684, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.885812520980835, "num_tokens": 625286990.0, "step": 16390 }, { "epoch": 2.0851036763770514, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.358943939208984, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8846439123153687, "num_tokens": 625326579.0, "step": 16391 }, { "epoch": 2.085230886655642, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.36279582977295, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8790745139122009, "num_tokens": 625367317.0, "step": 16392 }, { "epoch": 2.0853580969342325, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.421042442321777, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8686441779136658, "num_tokens": 625401047.0, "step": 16393 }, { "epoch": 2.085485307212823, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.403379440307617, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8595080375671387, "num_tokens": 625441878.0, "step": 16394 }, { "epoch": 2.0856125174914135, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.360175132751465, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.872088611125946, "num_tokens": 625477007.0, "step": 16395 }, { "epoch": 2.0857397277700036, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.364831924438477, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8858847618103027, "num_tokens": 625519908.0, "step": 16396 }, { "epoch": 2.085866938048594, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.432857513427734, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8767563104629517, "num_tokens": 625554946.0, "step": 16397 }, { "epoch": 2.0859941483271847, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.417527198791504, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8842777013778687, "num_tokens": 625592121.0, "step": 16398 }, { "epoch": 2.086121358605775, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.354101181030273, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8718437552452087, "num_tokens": 625630596.0, "step": 16399 }, { "epoch": 2.0862485688843657, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.482874870300293, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8562411069869995, "num_tokens": 625669535.0, "step": 16400 }, { "epoch": 2.0863757791629562, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.423702239990234, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8718941807746887, "num_tokens": 625710072.0, "step": 16401 }, { "epoch": 2.0865029894415468, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.376359939575195, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8810508251190186, "num_tokens": 625746328.0, "step": 16402 }, { "epoch": 2.0866301997201373, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.535284042358398, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8751866817474365, "num_tokens": 625787108.0, "step": 16403 }, { "epoch": 2.086757409998728, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39886474609375, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8778530359268188, "num_tokens": 625827887.0, "step": 16404 }, { "epoch": 2.0868846202773184, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.439282417297363, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8737173676490784, "num_tokens": 625863475.0, "step": 16405 }, { "epoch": 2.087011830555909, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.453770637512207, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.8893249034881592, "num_tokens": 625902369.0, "step": 16406 }, { "epoch": 2.0871390408344994, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442902565002441, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.84891277551651, "num_tokens": 625941054.0, "step": 16407 }, { "epoch": 2.08726625111309, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.479043006896973, "learning_rate": 1e-06, "loss": 0.4676, "mean_token_accuracy": 0.8556589484214783, "num_tokens": 625981113.0, "step": 16408 }, { "epoch": 2.0873934613916805, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.416777610778809, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8707598447799683, "num_tokens": 626014758.0, "step": 16409 }, { "epoch": 2.087520671670271, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.543899536132812, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8593612909317017, "num_tokens": 626054442.0, "step": 16410 }, { "epoch": 2.0876478819488615, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.484151840209961, "learning_rate": 1e-06, "loss": 0.4702, "mean_token_accuracy": 0.8560198545455933, "num_tokens": 626093993.0, "step": 16411 }, { "epoch": 2.087775092227452, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.469128608703613, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8561797142028809, "num_tokens": 626131242.0, "step": 16412 }, { "epoch": 2.0879023025060426, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.500340461730957, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8576657176017761, "num_tokens": 626162137.0, "step": 16413 }, { "epoch": 2.088029512784633, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.425945281982422, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8741996884346008, "num_tokens": 626201031.0, "step": 16414 }, { "epoch": 2.0881567230632236, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.415274620056152, "learning_rate": 1e-06, "loss": 0.3601, "mean_token_accuracy": 0.883685290813446, "num_tokens": 626243647.0, "step": 16415 }, { "epoch": 2.088283933341814, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.427769660949707, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8786883354187012, "num_tokens": 626281346.0, "step": 16416 }, { "epoch": 2.0884111436204047, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.542470932006836, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.882578432559967, "num_tokens": 626315964.0, "step": 16417 }, { "epoch": 2.0885383538989952, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.374968528747559, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.867159903049469, "num_tokens": 626367920.0, "step": 16418 }, { "epoch": 2.0886655641775858, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.384785652160645, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8703052997589111, "num_tokens": 626412123.0, "step": 16419 }, { "epoch": 2.0887927744561763, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.506509780883789, "learning_rate": 1e-06, "loss": 0.4838, "mean_token_accuracy": 0.846252977848053, "num_tokens": 626451560.0, "step": 16420 }, { "epoch": 2.0889199847347664, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.351670265197754, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.866788387298584, "num_tokens": 626489361.0, "step": 16421 }, { "epoch": 2.089047195013357, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.468670845031738, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8881044983863831, "num_tokens": 626527150.0, "step": 16422 }, { "epoch": 2.0891744052919474, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.37060260772705, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8657331466674805, "num_tokens": 626568110.0, "step": 16423 }, { "epoch": 2.089301615570538, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.520407676696777, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8745507001876831, "num_tokens": 626607215.0, "step": 16424 }, { "epoch": 2.0894288258491285, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.405223846435547, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8785707950592041, "num_tokens": 626644424.0, "step": 16425 }, { "epoch": 2.089556036127719, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.458759307861328, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8515042066574097, "num_tokens": 626689509.0, "step": 16426 }, { "epoch": 2.0896832464063095, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.449756622314453, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8627721071243286, "num_tokens": 626733866.0, "step": 16427 }, { "epoch": 2.0898104566849, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.44605541229248, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8815354704856873, "num_tokens": 626767465.0, "step": 16428 }, { "epoch": 2.0899376669634906, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.42175006866455, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.859320878982544, "num_tokens": 626806049.0, "step": 16429 }, { "epoch": 2.090064877242081, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.448493957519531, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8603812456130981, "num_tokens": 626848917.0, "step": 16430 }, { "epoch": 2.0901920875206716, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.476728439331055, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8597115874290466, "num_tokens": 626884442.0, "step": 16431 }, { "epoch": 2.090319297799262, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 13.409642219543457, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8710511326789856, "num_tokens": 626922517.0, "step": 16432 }, { "epoch": 2.0904465080778527, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.412308692932129, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8707020878791809, "num_tokens": 626962880.0, "step": 16433 }, { "epoch": 2.0905737183564432, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.465322494506836, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8796364665031433, "num_tokens": 627008115.0, "step": 16434 }, { "epoch": 2.0907009286350338, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.320558547973633, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8841524124145508, "num_tokens": 627043935.0, "step": 16435 }, { "epoch": 2.0908281389136243, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.463637351989746, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8739356994628906, "num_tokens": 627078495.0, "step": 16436 }, { "epoch": 2.090955349192215, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.412426948547363, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8766679763793945, "num_tokens": 627116652.0, "step": 16437 }, { "epoch": 2.0910825594708053, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.445308685302734, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8812528848648071, "num_tokens": 627152326.0, "step": 16438 }, { "epoch": 2.091209769749396, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.458637237548828, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8817843198776245, "num_tokens": 627189502.0, "step": 16439 }, { "epoch": 2.0913369800279864, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.440567970275879, "learning_rate": 1e-06, "loss": 0.352, "mean_token_accuracy": 0.8880541324615479, "num_tokens": 627225438.0, "step": 16440 }, { "epoch": 2.091464190306577, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.506464004516602, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8694552779197693, "num_tokens": 627258154.0, "step": 16441 }, { "epoch": 2.0915914005851675, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.425649642944336, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8652359247207642, "num_tokens": 627301804.0, "step": 16442 }, { "epoch": 2.091718610863758, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.395462036132812, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8762997388839722, "num_tokens": 627344301.0, "step": 16443 }, { "epoch": 2.0918458211423485, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.464704513549805, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8807923793792725, "num_tokens": 627381068.0, "step": 16444 }, { "epoch": 2.0919730314209386, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.440749168395996, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8666608333587646, "num_tokens": 627423410.0, "step": 16445 }, { "epoch": 2.092100241699529, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.34802532196045, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8824613094329834, "num_tokens": 627456593.0, "step": 16446 }, { "epoch": 2.0922274519781197, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.463446617126465, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8749488592147827, "num_tokens": 627493742.0, "step": 16447 }, { "epoch": 2.09235466225671, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.374178886413574, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8797149658203125, "num_tokens": 627529809.0, "step": 16448 }, { "epoch": 2.0924818725353007, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.4279203414917, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8765007257461548, "num_tokens": 627570693.0, "step": 16449 }, { "epoch": 2.0926090828138912, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.441047668457031, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8801297545433044, "num_tokens": 627607608.0, "step": 16450 }, { "epoch": 2.0927362930924818, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.456286430358887, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8717418909072876, "num_tokens": 627642102.0, "step": 16451 }, { "epoch": 2.0928635033710723, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.448807716369629, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8669530749320984, "num_tokens": 627678838.0, "step": 16452 }, { "epoch": 2.092990713649663, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.470364570617676, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8776038885116577, "num_tokens": 627714952.0, "step": 16453 }, { "epoch": 2.0931179239282534, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.371758460998535, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8632276654243469, "num_tokens": 627754710.0, "step": 16454 }, { "epoch": 2.093245134206844, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.472516059875488, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8779581785202026, "num_tokens": 627791504.0, "step": 16455 }, { "epoch": 2.0933723444854344, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.418644905090332, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8868299722671509, "num_tokens": 627829683.0, "step": 16456 }, { "epoch": 2.093499554764025, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.471266746520996, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8734428882598877, "num_tokens": 627869826.0, "step": 16457 }, { "epoch": 2.0936267650426155, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.455672264099121, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8698396682739258, "num_tokens": 627904165.0, "step": 16458 }, { "epoch": 2.093753975321206, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.4312105178833, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.878352165222168, "num_tokens": 627936559.0, "step": 16459 }, { "epoch": 2.0938811855997965, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.421920776367188, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8781085014343262, "num_tokens": 627971742.0, "step": 16460 }, { "epoch": 2.094008395878387, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.448918342590332, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8834185600280762, "num_tokens": 628013104.0, "step": 16461 }, { "epoch": 2.0941356061569776, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.487236022949219, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8712461590766907, "num_tokens": 628049378.0, "step": 16462 }, { "epoch": 2.094262816435568, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.413799285888672, "learning_rate": 1e-06, "loss": 0.3428, "mean_token_accuracy": 0.8910373449325562, "num_tokens": 628087213.0, "step": 16463 }, { "epoch": 2.0943900267141586, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.428973197937012, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8784183263778687, "num_tokens": 628122060.0, "step": 16464 }, { "epoch": 2.094517236992749, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.454362869262695, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8767035007476807, "num_tokens": 628166554.0, "step": 16465 }, { "epoch": 2.0946444472713397, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.434816360473633, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8721972703933716, "num_tokens": 628199953.0, "step": 16466 }, { "epoch": 2.09477165754993, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.467727661132812, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.876700758934021, "num_tokens": 628237044.0, "step": 16467 }, { "epoch": 2.0948988678285207, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.461377143859863, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8758671283721924, "num_tokens": 628275773.0, "step": 16468 }, { "epoch": 2.095026078107111, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.420092582702637, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8594406843185425, "num_tokens": 628314549.0, "step": 16469 }, { "epoch": 2.0951532883857014, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.463372230529785, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8697309494018555, "num_tokens": 628349417.0, "step": 16470 }, { "epoch": 2.095280498664292, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.482633590698242, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8778665065765381, "num_tokens": 628384435.0, "step": 16471 }, { "epoch": 2.0954077089428824, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.431429862976074, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8609330058097839, "num_tokens": 628424124.0, "step": 16472 }, { "epoch": 2.095534919221473, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.435750961303711, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8761271238327026, "num_tokens": 628462412.0, "step": 16473 }, { "epoch": 2.0956621295000635, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.484147071838379, "learning_rate": 1e-06, "loss": 0.4555, "mean_token_accuracy": 0.8560792207717896, "num_tokens": 628499066.0, "step": 16474 }, { "epoch": 2.095789339778654, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.440192222595215, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8878598213195801, "num_tokens": 628532379.0, "step": 16475 }, { "epoch": 2.0959165500572445, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.350255966186523, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8834055662155151, "num_tokens": 628570497.0, "step": 16476 }, { "epoch": 2.096043760335835, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.399797439575195, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8786760568618774, "num_tokens": 628608298.0, "step": 16477 }, { "epoch": 2.0961709706144256, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.44517707824707, "learning_rate": 1e-06, "loss": 0.3389, "mean_token_accuracy": 0.8920660614967346, "num_tokens": 628641217.0, "step": 16478 }, { "epoch": 2.096298180893016, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442529678344727, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8745349645614624, "num_tokens": 628682480.0, "step": 16479 }, { "epoch": 2.0964253911716066, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.393056869506836, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8719615340232849, "num_tokens": 628718359.0, "step": 16480 }, { "epoch": 2.096552601450197, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.47504997253418, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8808962106704712, "num_tokens": 628752214.0, "step": 16481 }, { "epoch": 2.0966798117287877, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.366985321044922, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8871756792068481, "num_tokens": 628787882.0, "step": 16482 }, { "epoch": 2.0968070220073782, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.455212593078613, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8707689642906189, "num_tokens": 628824910.0, "step": 16483 }, { "epoch": 2.0969342322859688, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.396360397338867, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.885055422782898, "num_tokens": 628860657.0, "step": 16484 }, { "epoch": 2.0970614425645593, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.421135902404785, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.858301043510437, "num_tokens": 628900619.0, "step": 16485 }, { "epoch": 2.09718865284315, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.463491439819336, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8859682679176331, "num_tokens": 628944304.0, "step": 16486 }, { "epoch": 2.0973158631217403, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.477684020996094, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8781453371047974, "num_tokens": 628980490.0, "step": 16487 }, { "epoch": 2.097443073400331, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.45338249206543, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8776462078094482, "num_tokens": 629011924.0, "step": 16488 }, { "epoch": 2.0975702836789214, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.46158504486084, "learning_rate": 1e-06, "loss": 0.4787, "mean_token_accuracy": 0.8495585918426514, "num_tokens": 629049514.0, "step": 16489 }, { "epoch": 2.097697493957512, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.467626571655273, "learning_rate": 1e-06, "loss": 0.3416, "mean_token_accuracy": 0.8911218643188477, "num_tokens": 629085374.0, "step": 16490 }, { "epoch": 2.0978247042361025, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.454602241516113, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.870294451713562, "num_tokens": 629126307.0, "step": 16491 }, { "epoch": 2.097951914514693, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.476395606994629, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8736469745635986, "num_tokens": 629159693.0, "step": 16492 }, { "epoch": 2.0980791247932835, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.388731956481934, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8792953491210938, "num_tokens": 629198544.0, "step": 16493 }, { "epoch": 2.0982063350718736, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.467668533325195, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.865714430809021, "num_tokens": 629240956.0, "step": 16494 }, { "epoch": 2.098333545350464, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.433883666992188, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8756104111671448, "num_tokens": 629276196.0, "step": 16495 }, { "epoch": 2.0984607556290547, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.510409355163574, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8619889616966248, "num_tokens": 629311398.0, "step": 16496 }, { "epoch": 2.098587965907645, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.447041511535645, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8691039085388184, "num_tokens": 629349488.0, "step": 16497 }, { "epoch": 2.0987151761862357, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.45173454284668, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8866859674453735, "num_tokens": 629388023.0, "step": 16498 }, { "epoch": 2.0988423864648262, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.382976531982422, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8728770613670349, "num_tokens": 629432140.0, "step": 16499 }, { "epoch": 2.0989695967434168, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.409957885742188, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8599010109901428, "num_tokens": 629480690.0, "step": 16500 }, { "epoch": 2.0990968070220073, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.41177749633789, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.890937089920044, "num_tokens": 629522500.0, "step": 16501 }, { "epoch": 2.099224017300598, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.346945762634277, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8752369284629822, "num_tokens": 629563450.0, "step": 16502 }, { "epoch": 2.0993512275791884, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.477396965026855, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8840135335922241, "num_tokens": 629603753.0, "step": 16503 }, { "epoch": 2.099478437857779, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.375085830688477, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8772276639938354, "num_tokens": 629637828.0, "step": 16504 }, { "epoch": 2.0996056481363694, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.414340019226074, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8816629648208618, "num_tokens": 629674636.0, "step": 16505 }, { "epoch": 2.09973285841496, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.37585735321045, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8644115924835205, "num_tokens": 629717046.0, "step": 16506 }, { "epoch": 2.0998600686935505, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442830085754395, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8839777112007141, "num_tokens": 629750331.0, "step": 16507 }, { "epoch": 2.099987278972141, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.389827728271484, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8556278944015503, "num_tokens": 629790095.0, "step": 16508 }, { "epoch": 2.1001144892507315, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.375617980957031, "learning_rate": 1e-06, "loss": 0.341, "mean_token_accuracy": 0.887893557548523, "num_tokens": 629823035.0, "step": 16509 }, { "epoch": 2.100241699529322, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.411908149719238, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8689878582954407, "num_tokens": 629857096.0, "step": 16510 }, { "epoch": 2.1003689098079126, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.455904006958008, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8728792667388916, "num_tokens": 629893448.0, "step": 16511 }, { "epoch": 2.100496120086503, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.400615692138672, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8709678649902344, "num_tokens": 629931301.0, "step": 16512 }, { "epoch": 2.1006233303650936, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.425210952758789, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8824201822280884, "num_tokens": 629963481.0, "step": 16513 }, { "epoch": 2.100750540643684, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.479717254638672, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8694766759872437, "num_tokens": 630002739.0, "step": 16514 }, { "epoch": 2.1008777509222747, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.433762550354004, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8670094013214111, "num_tokens": 630041206.0, "step": 16515 }, { "epoch": 2.101004961200865, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.454033851623535, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8600888252258301, "num_tokens": 630083811.0, "step": 16516 }, { "epoch": 2.1011321714794557, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.499093055725098, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.872083842754364, "num_tokens": 630119823.0, "step": 16517 }, { "epoch": 2.1012593817580463, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.448527336120605, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.873799741268158, "num_tokens": 630163807.0, "step": 16518 }, { "epoch": 2.1013865920366364, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.433601379394531, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8694015741348267, "num_tokens": 630203038.0, "step": 16519 }, { "epoch": 2.101513802315227, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.431565284729004, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8777758479118347, "num_tokens": 630238505.0, "step": 16520 }, { "epoch": 2.1016410125938174, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.398018836975098, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8728561401367188, "num_tokens": 630275279.0, "step": 16521 }, { "epoch": 2.101768222872408, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.416112899780273, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.870759129524231, "num_tokens": 630317288.0, "step": 16522 }, { "epoch": 2.1018954331509985, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.4879150390625, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8745138049125671, "num_tokens": 630352786.0, "step": 16523 }, { "epoch": 2.102022643429589, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.37772274017334, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8747245669364929, "num_tokens": 630391138.0, "step": 16524 }, { "epoch": 2.1021498537081795, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.470929145812988, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8627725839614868, "num_tokens": 630426567.0, "step": 16525 }, { "epoch": 2.10227706398677, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.521082878112793, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8632992506027222, "num_tokens": 630467308.0, "step": 16526 }, { "epoch": 2.1024042742653606, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.373011589050293, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8847202062606812, "num_tokens": 630503333.0, "step": 16527 }, { "epoch": 2.102531484543951, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.490610122680664, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.869012713432312, "num_tokens": 630536838.0, "step": 16528 }, { "epoch": 2.1026586948225416, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.424799919128418, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.886806845664978, "num_tokens": 630570833.0, "step": 16529 }, { "epoch": 2.102785905101132, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.436917304992676, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8823283314704895, "num_tokens": 630615043.0, "step": 16530 }, { "epoch": 2.1029131153797227, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.466998100280762, "learning_rate": 1e-06, "loss": 0.3367, "mean_token_accuracy": 0.8958389759063721, "num_tokens": 630651368.0, "step": 16531 }, { "epoch": 2.1030403256583132, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.494839668273926, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8669580817222595, "num_tokens": 630690341.0, "step": 16532 }, { "epoch": 2.1031675359369038, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.38836669921875, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.871303379535675, "num_tokens": 630727168.0, "step": 16533 }, { "epoch": 2.1032947462154943, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.428494453430176, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8785943984985352, "num_tokens": 630766737.0, "step": 16534 }, { "epoch": 2.103421956494085, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.424239158630371, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8715678453445435, "num_tokens": 630803721.0, "step": 16535 }, { "epoch": 2.1035491667726753, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.36841869354248, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8805400729179382, "num_tokens": 630843697.0, "step": 16536 }, { "epoch": 2.103676377051266, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.434261322021484, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8746138215065002, "num_tokens": 630874325.0, "step": 16537 }, { "epoch": 2.1038035873298564, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.438920021057129, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8700539469718933, "num_tokens": 630911456.0, "step": 16538 }, { "epoch": 2.103930797608447, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.423967361450195, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8722867369651794, "num_tokens": 630954418.0, "step": 16539 }, { "epoch": 2.1040580078870375, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.398603439331055, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8735575079917908, "num_tokens": 630991826.0, "step": 16540 }, { "epoch": 2.104185218165628, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.454574584960938, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8741543292999268, "num_tokens": 631025831.0, "step": 16541 }, { "epoch": 2.1043124284442185, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.436684608459473, "learning_rate": 1e-06, "loss": 0.4532, "mean_token_accuracy": 0.85892254114151, "num_tokens": 631062896.0, "step": 16542 }, { "epoch": 2.1044396387228086, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.436813354492188, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8799856901168823, "num_tokens": 631099291.0, "step": 16543 }, { "epoch": 2.104566849001399, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.489657402038574, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8656394481658936, "num_tokens": 631141570.0, "step": 16544 }, { "epoch": 2.1046940592799896, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.4252290725708, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8662346601486206, "num_tokens": 631178882.0, "step": 16545 }, { "epoch": 2.10482126955858, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.45724868774414, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8703140616416931, "num_tokens": 631218059.0, "step": 16546 }, { "epoch": 2.1049484798371707, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.501253128051758, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8749375343322754, "num_tokens": 631252827.0, "step": 16547 }, { "epoch": 2.1050756901157612, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.427916526794434, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8811420202255249, "num_tokens": 631294370.0, "step": 16548 }, { "epoch": 2.1052029003943518, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.507948875427246, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8706170320510864, "num_tokens": 631331091.0, "step": 16549 }, { "epoch": 2.1053301106729423, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.472787857055664, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8769931197166443, "num_tokens": 631367285.0, "step": 16550 }, { "epoch": 2.105457320951533, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.432300567626953, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8595603704452515, "num_tokens": 631402471.0, "step": 16551 }, { "epoch": 2.1055845312301233, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.434608459472656, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8733890056610107, "num_tokens": 631439706.0, "step": 16552 }, { "epoch": 2.105711741508714, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.39165210723877, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8786402344703674, "num_tokens": 631471437.0, "step": 16553 }, { "epoch": 2.1058389517873044, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.42867374420166, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8710541129112244, "num_tokens": 631515028.0, "step": 16554 }, { "epoch": 2.105966162065895, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.416017532348633, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.869248628616333, "num_tokens": 631552946.0, "step": 16555 }, { "epoch": 2.1060933723444855, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.424524307250977, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8778651356697083, "num_tokens": 631596045.0, "step": 16556 }, { "epoch": 2.106220582623076, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.47179889678955, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8811384439468384, "num_tokens": 631630233.0, "step": 16557 }, { "epoch": 2.1063477929016665, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.432889938354492, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8801612854003906, "num_tokens": 631666665.0, "step": 16558 }, { "epoch": 2.106475003180257, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.481557846069336, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8703912496566772, "num_tokens": 631709536.0, "step": 16559 }, { "epoch": 2.1066022134588476, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.491240501403809, "learning_rate": 1e-06, "loss": 0.3477, "mean_token_accuracy": 0.8880020380020142, "num_tokens": 631744333.0, "step": 16560 }, { "epoch": 2.106729423737438, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.47216796875, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8597819805145264, "num_tokens": 631783522.0, "step": 16561 }, { "epoch": 2.1068566340160286, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.522211074829102, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8586738705635071, "num_tokens": 631822579.0, "step": 16562 }, { "epoch": 2.106983844294619, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.490715980529785, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8808888792991638, "num_tokens": 631862642.0, "step": 16563 }, { "epoch": 2.1071110545732097, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.395929336547852, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8841726779937744, "num_tokens": 631903161.0, "step": 16564 }, { "epoch": 2.1072382648518, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.519563674926758, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8766037821769714, "num_tokens": 631938149.0, "step": 16565 }, { "epoch": 2.1073654751303907, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.458785057067871, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8757646083831787, "num_tokens": 631974717.0, "step": 16566 }, { "epoch": 2.107492685408981, "ewc_loss": 0.03466796875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 13.494609832763672, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8776226043701172, "num_tokens": 632011147.0, "step": 16567 }, { "epoch": 2.1076198956875714, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.45101547241211, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.879976749420166, "num_tokens": 632051915.0, "step": 16568 }, { "epoch": 2.107747105966162, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.499817848205566, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8649621605873108, "num_tokens": 632083047.0, "step": 16569 }, { "epoch": 2.1078743162447524, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.40804672241211, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8860988020896912, "num_tokens": 632119250.0, "step": 16570 }, { "epoch": 2.108001526523343, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.5023775100708, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8806642889976501, "num_tokens": 632152696.0, "step": 16571 }, { "epoch": 2.1081287368019335, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.456624031066895, "learning_rate": 1e-06, "loss": 0.3379, "mean_token_accuracy": 0.8940986394882202, "num_tokens": 632188167.0, "step": 16572 }, { "epoch": 2.108255947080524, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.48132038116455, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8724386692047119, "num_tokens": 632229567.0, "step": 16573 }, { "epoch": 2.1083831573591145, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.429943084716797, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.882936954498291, "num_tokens": 632266870.0, "step": 16574 }, { "epoch": 2.108510367637705, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.480199813842773, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.873045027256012, "num_tokens": 632303844.0, "step": 16575 }, { "epoch": 2.1086375779162956, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.49701976776123, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8755267262458801, "num_tokens": 632339343.0, "step": 16576 }, { "epoch": 2.108764788194886, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.468843460083008, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8790333867073059, "num_tokens": 632385516.0, "step": 16577 }, { "epoch": 2.1088919984734766, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.440671920776367, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8784152865409851, "num_tokens": 632420664.0, "step": 16578 }, { "epoch": 2.109019208752067, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.513866424560547, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8792505264282227, "num_tokens": 632458082.0, "step": 16579 }, { "epoch": 2.1091464190306577, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.497896194458008, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8629767894744873, "num_tokens": 632495963.0, "step": 16580 }, { "epoch": 2.109273629309248, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.47485065460205, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8751909136772156, "num_tokens": 632529968.0, "step": 16581 }, { "epoch": 2.1094008395878387, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442977905273438, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8824384808540344, "num_tokens": 632567527.0, "step": 16582 }, { "epoch": 2.1095280498664293, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.490828514099121, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.873898983001709, "num_tokens": 632606192.0, "step": 16583 }, { "epoch": 2.10965526014502, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.441044807434082, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8850157856941223, "num_tokens": 632645940.0, "step": 16584 }, { "epoch": 2.1097824704236103, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.429149627685547, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8825442790985107, "num_tokens": 632689385.0, "step": 16585 }, { "epoch": 2.109909680702201, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.528797149658203, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8823012709617615, "num_tokens": 632720178.0, "step": 16586 }, { "epoch": 2.1100368909807914, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.485315322875977, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8673224449157715, "num_tokens": 632763249.0, "step": 16587 }, { "epoch": 2.110164101259382, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.485137939453125, "learning_rate": 1e-06, "loss": 0.454, "mean_token_accuracy": 0.8561537861824036, "num_tokens": 632803293.0, "step": 16588 }, { "epoch": 2.1102913115379724, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.471951484680176, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8652690649032593, "num_tokens": 632838731.0, "step": 16589 }, { "epoch": 2.110418521816563, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.463722229003906, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8651182651519775, "num_tokens": 632875044.0, "step": 16590 }, { "epoch": 2.1105457320951535, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.472890853881836, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8540268540382385, "num_tokens": 632911330.0, "step": 16591 }, { "epoch": 2.1106729423737436, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.514740943908691, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8638678789138794, "num_tokens": 632952122.0, "step": 16592 }, { "epoch": 2.110800152652334, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.475045204162598, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8565686941146851, "num_tokens": 632985602.0, "step": 16593 }, { "epoch": 2.1109273629309246, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.466251373291016, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8751767873764038, "num_tokens": 633023792.0, "step": 16594 }, { "epoch": 2.111054573209515, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.459015846252441, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8753554821014404, "num_tokens": 633065457.0, "step": 16595 }, { "epoch": 2.1111817834881057, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.477824211120605, "learning_rate": 1e-06, "loss": 0.3582, "mean_token_accuracy": 0.883237361907959, "num_tokens": 633101402.0, "step": 16596 }, { "epoch": 2.1113089937666962, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.467594146728516, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8649582862854004, "num_tokens": 633147002.0, "step": 16597 }, { "epoch": 2.1114362040452868, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.518375396728516, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8735409379005432, "num_tokens": 633184209.0, "step": 16598 }, { "epoch": 2.1115634143238773, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.4656400680542, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8784893751144409, "num_tokens": 633224525.0, "step": 16599 }, { "epoch": 2.111690624602468, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.615453720092773, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.889159619808197, "num_tokens": 633253171.0, "step": 16600 }, { "epoch": 2.1118178348810583, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.49585247039795, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8710307478904724, "num_tokens": 633294326.0, "step": 16601 }, { "epoch": 2.111945045159649, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.476153373718262, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8868883848190308, "num_tokens": 633327236.0, "step": 16602 }, { "epoch": 2.1120722554382394, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.445298194885254, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8836668729782104, "num_tokens": 633368593.0, "step": 16603 }, { "epoch": 2.11219946571683, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.441617965698242, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.853312075138092, "num_tokens": 633402598.0, "step": 16604 }, { "epoch": 2.1123266759954205, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.504555702209473, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8690589070320129, "num_tokens": 633437069.0, "step": 16605 }, { "epoch": 2.112453886274011, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.424299240112305, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8767008781433105, "num_tokens": 633479752.0, "step": 16606 }, { "epoch": 2.1125810965526015, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.476512908935547, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8707437515258789, "num_tokens": 633516949.0, "step": 16607 }, { "epoch": 2.112708306831192, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.531771659851074, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8782182931900024, "num_tokens": 633547511.0, "step": 16608 }, { "epoch": 2.1128355171097826, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.436022758483887, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8763927221298218, "num_tokens": 633583940.0, "step": 16609 }, { "epoch": 2.112962727388373, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.480170249938965, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8636819124221802, "num_tokens": 633626245.0, "step": 16610 }, { "epoch": 2.1130899376669636, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.48969554901123, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.877444863319397, "num_tokens": 633662889.0, "step": 16611 }, { "epoch": 2.113217147945554, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.442468643188477, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8698999881744385, "num_tokens": 633701461.0, "step": 16612 }, { "epoch": 2.1133443582241447, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.44288444519043, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8758488893508911, "num_tokens": 633741302.0, "step": 16613 }, { "epoch": 2.113471568502735, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.46821403503418, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8792600631713867, "num_tokens": 633780337.0, "step": 16614 }, { "epoch": 2.1135987787813257, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.441319465637207, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8809472918510437, "num_tokens": 633814450.0, "step": 16615 }, { "epoch": 2.1137259890599163, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.513744354248047, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8669236302375793, "num_tokens": 633856252.0, "step": 16616 }, { "epoch": 2.1138531993385064, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.411203384399414, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8628747463226318, "num_tokens": 633895235.0, "step": 16617 }, { "epoch": 2.113980409617097, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.437264442443848, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8760508894920349, "num_tokens": 633938948.0, "step": 16618 }, { "epoch": 2.1141076198956874, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.430723190307617, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8807901740074158, "num_tokens": 633973393.0, "step": 16619 }, { "epoch": 2.114234830174278, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.43432331085205, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8788143396377563, "num_tokens": 634006714.0, "step": 16620 }, { "epoch": 2.1143620404528685, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.470641136169434, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.880120038986206, "num_tokens": 634046104.0, "step": 16621 }, { "epoch": 2.114489250731459, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.466206550598145, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8657982349395752, "num_tokens": 634086401.0, "step": 16622 }, { "epoch": 2.1146164610100495, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.508281707763672, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8657371997833252, "num_tokens": 634125479.0, "step": 16623 }, { "epoch": 2.11474367128864, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.446467399597168, "learning_rate": 1e-06, "loss": 0.4569, "mean_token_accuracy": 0.855671763420105, "num_tokens": 634163873.0, "step": 16624 }, { "epoch": 2.1148708815672306, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.478124618530273, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8713376522064209, "num_tokens": 634200533.0, "step": 16625 }, { "epoch": 2.114998091845821, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.49478816986084, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8666610717773438, "num_tokens": 634235891.0, "step": 16626 }, { "epoch": 2.1151253021244116, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.527095794677734, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8726892471313477, "num_tokens": 634266748.0, "step": 16627 }, { "epoch": 2.115252512403002, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.497714042663574, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8627893924713135, "num_tokens": 634302603.0, "step": 16628 }, { "epoch": 2.1153797226815927, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.465080261230469, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8756004571914673, "num_tokens": 634345471.0, "step": 16629 }, { "epoch": 2.115506932960183, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.498824119567871, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8723770380020142, "num_tokens": 634383902.0, "step": 16630 }, { "epoch": 2.1156341432387737, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.449233055114746, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8674668669700623, "num_tokens": 634421082.0, "step": 16631 }, { "epoch": 2.1157613535173643, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.515162467956543, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8680953979492188, "num_tokens": 634456490.0, "step": 16632 }, { "epoch": 2.115888563795955, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.507039070129395, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8556219935417175, "num_tokens": 634498252.0, "step": 16633 }, { "epoch": 2.1160157740745453, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.431593894958496, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.88149493932724, "num_tokens": 634540769.0, "step": 16634 }, { "epoch": 2.116142984353136, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.457809448242188, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8807698488235474, "num_tokens": 634573807.0, "step": 16635 }, { "epoch": 2.1162701946317264, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.467371940612793, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8690711259841919, "num_tokens": 634609542.0, "step": 16636 }, { "epoch": 2.116397404910317, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.421615600585938, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8814866542816162, "num_tokens": 634648641.0, "step": 16637 }, { "epoch": 2.1165246151889074, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.482701301574707, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8718230724334717, "num_tokens": 634685531.0, "step": 16638 }, { "epoch": 2.116651825467498, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.477765083312988, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.875119686126709, "num_tokens": 634717239.0, "step": 16639 }, { "epoch": 2.116779035746088, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.472787857055664, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8703829646110535, "num_tokens": 634750058.0, "step": 16640 }, { "epoch": 2.1169062460246786, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.458211898803711, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8668738603591919, "num_tokens": 634793480.0, "step": 16641 }, { "epoch": 2.117033456303269, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.473995208740234, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8639335036277771, "num_tokens": 634829923.0, "step": 16642 }, { "epoch": 2.1171606665818596, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.519688606262207, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8687242269515991, "num_tokens": 634868499.0, "step": 16643 }, { "epoch": 2.11728787686045, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.498553276062012, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8811308145523071, "num_tokens": 634905922.0, "step": 16644 }, { "epoch": 2.1174150871390407, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.468440055847168, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8742255568504333, "num_tokens": 634946190.0, "step": 16645 }, { "epoch": 2.1175422974176312, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.415339469909668, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8856643438339233, "num_tokens": 634981840.0, "step": 16646 }, { "epoch": 2.1176695076962218, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.498605728149414, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8715803623199463, "num_tokens": 635024755.0, "step": 16647 }, { "epoch": 2.1177967179748123, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.476447105407715, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8770043253898621, "num_tokens": 635063076.0, "step": 16648 }, { "epoch": 2.117923928253403, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.415637969970703, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8859678506851196, "num_tokens": 635103052.0, "step": 16649 }, { "epoch": 2.1180511385319933, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.427766799926758, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8828637599945068, "num_tokens": 635140917.0, "step": 16650 }, { "epoch": 2.118178348810584, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.448580741882324, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8778916597366333, "num_tokens": 635182049.0, "step": 16651 }, { "epoch": 2.1183055590891744, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.43096923828125, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.868750810623169, "num_tokens": 635217613.0, "step": 16652 }, { "epoch": 2.118432769367765, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.42361831665039, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8830944299697876, "num_tokens": 635252716.0, "step": 16653 }, { "epoch": 2.1185599796463555, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.457054138183594, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8771049976348877, "num_tokens": 635288919.0, "step": 16654 }, { "epoch": 2.118687189924946, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.496389389038086, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8687152862548828, "num_tokens": 635329781.0, "step": 16655 }, { "epoch": 2.1188144002035365, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.507436752319336, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8832563161849976, "num_tokens": 635363844.0, "step": 16656 }, { "epoch": 2.118941610482127, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.455156326293945, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8704416155815125, "num_tokens": 635397568.0, "step": 16657 }, { "epoch": 2.1190688207607176, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.473012924194336, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8579808473587036, "num_tokens": 635432903.0, "step": 16658 }, { "epoch": 2.119196031039308, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.441676139831543, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.876500129699707, "num_tokens": 635473388.0, "step": 16659 }, { "epoch": 2.1193232413178986, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.483656883239746, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8777225613594055, "num_tokens": 635512966.0, "step": 16660 }, { "epoch": 2.119450451596489, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.484559059143066, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.860765278339386, "num_tokens": 635547814.0, "step": 16661 }, { "epoch": 2.1195776618750797, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.442365646362305, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8645398616790771, "num_tokens": 635583120.0, "step": 16662 }, { "epoch": 2.11970487215367, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.427635192871094, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8690235614776611, "num_tokens": 635621531.0, "step": 16663 }, { "epoch": 2.1198320824322607, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.458846092224121, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8635459542274475, "num_tokens": 635659641.0, "step": 16664 }, { "epoch": 2.119959292710851, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.415172576904297, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8687548041343689, "num_tokens": 635697953.0, "step": 16665 }, { "epoch": 2.1200865029894413, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.460258483886719, "learning_rate": 1e-06, "loss": 0.4716, "mean_token_accuracy": 0.8484758138656616, "num_tokens": 635742519.0, "step": 16666 }, { "epoch": 2.120213713268032, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.47511100769043, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.881314754486084, "num_tokens": 635777786.0, "step": 16667 }, { "epoch": 2.1203409235466224, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.491362571716309, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8764299154281616, "num_tokens": 635814490.0, "step": 16668 }, { "epoch": 2.120468133825213, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.479979515075684, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8604199290275574, "num_tokens": 635858351.0, "step": 16669 }, { "epoch": 2.1205953441038035, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.48473834991455, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8712084293365479, "num_tokens": 635895585.0, "step": 16670 }, { "epoch": 2.120722554382394, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.497871398925781, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8810527324676514, "num_tokens": 635933591.0, "step": 16671 }, { "epoch": 2.1208497646609845, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.485149383544922, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8708548545837402, "num_tokens": 635972040.0, "step": 16672 }, { "epoch": 2.120976974939575, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.487719535827637, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8766449689865112, "num_tokens": 636013287.0, "step": 16673 }, { "epoch": 2.1211041852181656, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.480121612548828, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8687112927436829, "num_tokens": 636046137.0, "step": 16674 }, { "epoch": 2.121231395496756, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.524665832519531, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8626934289932251, "num_tokens": 636088870.0, "step": 16675 }, { "epoch": 2.1213586057753466, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.473455429077148, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.883385181427002, "num_tokens": 636130519.0, "step": 16676 }, { "epoch": 2.121485816053937, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.44816780090332, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8763877153396606, "num_tokens": 636165641.0, "step": 16677 }, { "epoch": 2.1216130263325277, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.515776634216309, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8602772355079651, "num_tokens": 636203811.0, "step": 16678 }, { "epoch": 2.121740236611118, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.48702621459961, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.885041356086731, "num_tokens": 636239503.0, "step": 16679 }, { "epoch": 2.1218674468897087, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.456785202026367, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8766688108444214, "num_tokens": 636278994.0, "step": 16680 }, { "epoch": 2.1219946571682993, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.474676132202148, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8694980144500732, "num_tokens": 636318025.0, "step": 16681 }, { "epoch": 2.12212186744689, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.465883255004883, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8754057288169861, "num_tokens": 636357289.0, "step": 16682 }, { "epoch": 2.1222490777254803, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.480237007141113, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.858646810054779, "num_tokens": 636399067.0, "step": 16683 }, { "epoch": 2.122376288004071, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.378829956054688, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8787806034088135, "num_tokens": 636442418.0, "step": 16684 }, { "epoch": 2.1225034982826614, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.498708724975586, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8612265586853027, "num_tokens": 636484273.0, "step": 16685 }, { "epoch": 2.122630708561252, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.4252347946167, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8829736113548279, "num_tokens": 636525675.0, "step": 16686 }, { "epoch": 2.1227579188398424, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.541325569152832, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8729199171066284, "num_tokens": 636559652.0, "step": 16687 }, { "epoch": 2.122885129118433, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.370657920837402, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8912220001220703, "num_tokens": 636597900.0, "step": 16688 }, { "epoch": 2.1230123393970235, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.520339012145996, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.874817967414856, "num_tokens": 636635822.0, "step": 16689 }, { "epoch": 2.1231395496756136, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.392946243286133, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8636859655380249, "num_tokens": 636667275.0, "step": 16690 }, { "epoch": 2.123266759954204, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.521174430847168, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8758960962295532, "num_tokens": 636704723.0, "step": 16691 }, { "epoch": 2.1233939702327946, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.459453582763672, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.878139317035675, "num_tokens": 636748209.0, "step": 16692 }, { "epoch": 2.123521180511385, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.465215682983398, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8791376948356628, "num_tokens": 636785052.0, "step": 16693 }, { "epoch": 2.1236483907899757, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.521804809570312, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8863701820373535, "num_tokens": 636816073.0, "step": 16694 }, { "epoch": 2.123775601068566, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.461199760437012, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8802313208580017, "num_tokens": 636848963.0, "step": 16695 }, { "epoch": 2.1239028113471567, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.504034996032715, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8867738246917725, "num_tokens": 636880632.0, "step": 16696 }, { "epoch": 2.1240300216257473, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.444583892822266, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8606388568878174, "num_tokens": 636923597.0, "step": 16697 }, { "epoch": 2.124157231904338, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.489153861999512, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8681524395942688, "num_tokens": 636962250.0, "step": 16698 }, { "epoch": 2.1242844421829283, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.454903602600098, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8807988166809082, "num_tokens": 636997973.0, "step": 16699 }, { "epoch": 2.124411652461519, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.522626876831055, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8622904419898987, "num_tokens": 637037320.0, "step": 16700 }, { "epoch": 2.1245388627401094, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.482282638549805, "learning_rate": 1e-06, "loss": 0.4639, "mean_token_accuracy": 0.8523339033126831, "num_tokens": 637072267.0, "step": 16701 }, { "epoch": 2.1246660730187, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.486761093139648, "learning_rate": 1e-06, "loss": 0.3525, "mean_token_accuracy": 0.8882444500923157, "num_tokens": 637108782.0, "step": 16702 }, { "epoch": 2.1247932832972904, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.493749618530273, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8732361793518066, "num_tokens": 637147368.0, "step": 16703 }, { "epoch": 2.124920493575881, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.435494422912598, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8817567229270935, "num_tokens": 637181909.0, "step": 16704 }, { "epoch": 2.1250477038544715, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.507796287536621, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8733537793159485, "num_tokens": 637218498.0, "step": 16705 }, { "epoch": 2.125174914133062, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.493521690368652, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8750264644622803, "num_tokens": 637254825.0, "step": 16706 }, { "epoch": 2.1253021244116526, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.430078506469727, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8638571500778198, "num_tokens": 637301593.0, "step": 16707 }, { "epoch": 2.125429334690243, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.431596755981445, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8850544691085815, "num_tokens": 637341676.0, "step": 16708 }, { "epoch": 2.1255565449688336, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.453076362609863, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8784360885620117, "num_tokens": 637378840.0, "step": 16709 }, { "epoch": 2.125683755247424, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.508793830871582, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8570351004600525, "num_tokens": 637412711.0, "step": 16710 }, { "epoch": 2.1258109655260147, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.466703414916992, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.872110903263092, "num_tokens": 637448287.0, "step": 16711 }, { "epoch": 2.125938175804605, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.410653114318848, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8718054294586182, "num_tokens": 637487899.0, "step": 16712 }, { "epoch": 2.1260653860831957, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.467537879943848, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8790282011032104, "num_tokens": 637528161.0, "step": 16713 }, { "epoch": 2.1261925963617863, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.467522621154785, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8629705905914307, "num_tokens": 637565692.0, "step": 16714 }, { "epoch": 2.1263198066403763, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.467153549194336, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8610327243804932, "num_tokens": 637600368.0, "step": 16715 }, { "epoch": 2.126447016918967, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.50508975982666, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8709617853164673, "num_tokens": 637635732.0, "step": 16716 }, { "epoch": 2.1265742271975574, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.519338607788086, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8691400289535522, "num_tokens": 637677127.0, "step": 16717 }, { "epoch": 2.126701437476148, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.401022911071777, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8740782737731934, "num_tokens": 637715278.0, "step": 16718 }, { "epoch": 2.1268286477547385, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.482813835144043, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.876673698425293, "num_tokens": 637742937.0, "step": 16719 }, { "epoch": 2.126955858033329, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.476563453674316, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8755947351455688, "num_tokens": 637776761.0, "step": 16720 }, { "epoch": 2.1270830683119195, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.434772491455078, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8757559061050415, "num_tokens": 637812197.0, "step": 16721 }, { "epoch": 2.12721027859051, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.445879936218262, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8775405883789062, "num_tokens": 637850442.0, "step": 16722 }, { "epoch": 2.1273374888691006, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.45722484588623, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8747719526290894, "num_tokens": 637889352.0, "step": 16723 }, { "epoch": 2.127464699147691, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.533145904541016, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8564436435699463, "num_tokens": 637927195.0, "step": 16724 }, { "epoch": 2.1275919094262816, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.445147514343262, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8625073432922363, "num_tokens": 637969431.0, "step": 16725 }, { "epoch": 2.127719119704872, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.443550109863281, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8899630904197693, "num_tokens": 638013910.0, "step": 16726 }, { "epoch": 2.1278463299834627, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.49718952178955, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8637442588806152, "num_tokens": 638051848.0, "step": 16727 }, { "epoch": 2.127973540262053, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.46996021270752, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8738811016082764, "num_tokens": 638092192.0, "step": 16728 }, { "epoch": 2.1281007505406437, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.475451469421387, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8827807903289795, "num_tokens": 638134306.0, "step": 16729 }, { "epoch": 2.1282279608192343, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.429159164428711, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8597322702407837, "num_tokens": 638177988.0, "step": 16730 }, { "epoch": 2.128355171097825, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.461392402648926, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8848744630813599, "num_tokens": 638212237.0, "step": 16731 }, { "epoch": 2.1284823813764153, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.513656616210938, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8662431836128235, "num_tokens": 638252492.0, "step": 16732 }, { "epoch": 2.128609591655006, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.4507474899292, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8731637001037598, "num_tokens": 638294308.0, "step": 16733 }, { "epoch": 2.1287368019335964, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.474825859069824, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8766217827796936, "num_tokens": 638330178.0, "step": 16734 }, { "epoch": 2.128864012212187, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.49887466430664, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8534742593765259, "num_tokens": 638372731.0, "step": 16735 }, { "epoch": 2.1289912224907774, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.456229209899902, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8693286180496216, "num_tokens": 638407522.0, "step": 16736 }, { "epoch": 2.129118432769368, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.497308731079102, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8762023448944092, "num_tokens": 638443262.0, "step": 16737 }, { "epoch": 2.129245643047958, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.479077339172363, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8734331130981445, "num_tokens": 638482729.0, "step": 16738 }, { "epoch": 2.129372853326549, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.452265739440918, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8818187713623047, "num_tokens": 638520034.0, "step": 16739 }, { "epoch": 2.129500063605139, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.521267890930176, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.880963921546936, "num_tokens": 638558631.0, "step": 16740 }, { "epoch": 2.1296272738837296, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.453207969665527, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8847882747650146, "num_tokens": 638593565.0, "step": 16741 }, { "epoch": 2.12975448416232, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.509679794311523, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8791162967681885, "num_tokens": 638625952.0, "step": 16742 }, { "epoch": 2.1298816944409107, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.483250617980957, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8707267045974731, "num_tokens": 638662000.0, "step": 16743 }, { "epoch": 2.130008904719501, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.491106986999512, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8594976663589478, "num_tokens": 638706020.0, "step": 16744 }, { "epoch": 2.1301361149980917, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.480224609375, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8802395462989807, "num_tokens": 638745354.0, "step": 16745 }, { "epoch": 2.1302633252766823, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.428132057189941, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8678308725357056, "num_tokens": 638785497.0, "step": 16746 }, { "epoch": 2.130390535555273, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.537665367126465, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8531426191329956, "num_tokens": 638821650.0, "step": 16747 }, { "epoch": 2.1305177458338633, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.555855751037598, "learning_rate": 1e-06, "loss": 0.4399, "mean_token_accuracy": 0.8599321842193604, "num_tokens": 638855589.0, "step": 16748 }, { "epoch": 2.130644956112454, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.460990905761719, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8725329637527466, "num_tokens": 638895766.0, "step": 16749 }, { "epoch": 2.1307721663910444, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.479985237121582, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8638975620269775, "num_tokens": 638932622.0, "step": 16750 }, { "epoch": 2.130899376669635, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.506840705871582, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.863308846950531, "num_tokens": 638971588.0, "step": 16751 }, { "epoch": 2.1310265869482254, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.470479965209961, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8794273138046265, "num_tokens": 639009921.0, "step": 16752 }, { "epoch": 2.131153797226816, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.52074146270752, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8734880685806274, "num_tokens": 639048160.0, "step": 16753 }, { "epoch": 2.1312810075054065, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.474285125732422, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.8853141069412231, "num_tokens": 639080260.0, "step": 16754 }, { "epoch": 2.131408217783997, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.516876220703125, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8660751581192017, "num_tokens": 639118758.0, "step": 16755 }, { "epoch": 2.1315354280625876, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.522021293640137, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8743857145309448, "num_tokens": 639159528.0, "step": 16756 }, { "epoch": 2.131662638341178, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.49976921081543, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.877047598361969, "num_tokens": 639200394.0, "step": 16757 }, { "epoch": 2.1317898486197686, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.437270164489746, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8642264604568481, "num_tokens": 639240938.0, "step": 16758 }, { "epoch": 2.131917058898359, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.497496604919434, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8682044148445129, "num_tokens": 639278097.0, "step": 16759 }, { "epoch": 2.1320442691769497, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.448996543884277, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.872835636138916, "num_tokens": 639324512.0, "step": 16760 }, { "epoch": 2.13217147945554, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.50473690032959, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8784892559051514, "num_tokens": 639366833.0, "step": 16761 }, { "epoch": 2.1322986897341307, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.531933784484863, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8593062162399292, "num_tokens": 639405765.0, "step": 16762 }, { "epoch": 2.132425900012721, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.489602088928223, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8699312210083008, "num_tokens": 639439622.0, "step": 16763 }, { "epoch": 2.1325531102913113, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.488848686218262, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8838115334510803, "num_tokens": 639479943.0, "step": 16764 }, { "epoch": 2.132680320569902, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.45821475982666, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8676458597183228, "num_tokens": 639520487.0, "step": 16765 }, { "epoch": 2.1328075308484924, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.482034683227539, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8725675344467163, "num_tokens": 639558997.0, "step": 16766 }, { "epoch": 2.132934741127083, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.4553861618042, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8724355101585388, "num_tokens": 639594569.0, "step": 16767 }, { "epoch": 2.1330619514056735, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.5133638381958, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8872827291488647, "num_tokens": 639629827.0, "step": 16768 }, { "epoch": 2.133189161684264, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.499190330505371, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8677439093589783, "num_tokens": 639667634.0, "step": 16769 }, { "epoch": 2.1333163719628545, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.519491195678711, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8712084293365479, "num_tokens": 639700536.0, "step": 16770 }, { "epoch": 2.133443582241445, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.477251052856445, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8751816749572754, "num_tokens": 639744006.0, "step": 16771 }, { "epoch": 2.1335707925200356, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.506624221801758, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8784749507904053, "num_tokens": 639784950.0, "step": 16772 }, { "epoch": 2.133698002798626, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.520994186401367, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8865605592727661, "num_tokens": 639823567.0, "step": 16773 }, { "epoch": 2.1338252130772166, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.542909622192383, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8846970796585083, "num_tokens": 639863386.0, "step": 16774 }, { "epoch": 2.133952423355807, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.501212120056152, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8618319630622864, "num_tokens": 639901219.0, "step": 16775 }, { "epoch": 2.1340796336343977, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.594542503356934, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8568887710571289, "num_tokens": 639935240.0, "step": 16776 }, { "epoch": 2.134206843912988, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.50871467590332, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8653703331947327, "num_tokens": 639975370.0, "step": 16777 }, { "epoch": 2.1343340541915787, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.517258644104004, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.855171799659729, "num_tokens": 640020463.0, "step": 16778 }, { "epoch": 2.1344612644701693, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.555853843688965, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8677399158477783, "num_tokens": 640057142.0, "step": 16779 }, { "epoch": 2.13458847474876, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.506552696228027, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8691564798355103, "num_tokens": 640098879.0, "step": 16780 }, { "epoch": 2.1347156850273503, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.522175788879395, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8753975629806519, "num_tokens": 640136077.0, "step": 16781 }, { "epoch": 2.134842895305941, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.531390190124512, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8588414788246155, "num_tokens": 640175160.0, "step": 16782 }, { "epoch": 2.1349701055845314, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.485389709472656, "learning_rate": 1e-06, "loss": 0.3295, "mean_token_accuracy": 0.8972195386886597, "num_tokens": 640216670.0, "step": 16783 }, { "epoch": 2.135097315863122, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.515654563903809, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8707068562507629, "num_tokens": 640254872.0, "step": 16784 }, { "epoch": 2.1352245261417124, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.538639068603516, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8715067505836487, "num_tokens": 640289319.0, "step": 16785 }, { "epoch": 2.135351736420303, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.514049530029297, "learning_rate": 1e-06, "loss": 0.3524, "mean_token_accuracy": 0.8876933455467224, "num_tokens": 640323897.0, "step": 16786 }, { "epoch": 2.1354789466988935, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.431352615356445, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8785003423690796, "num_tokens": 640366203.0, "step": 16787 }, { "epoch": 2.1356061569774836, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.5143404006958, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.882773220539093, "num_tokens": 640408089.0, "step": 16788 }, { "epoch": 2.135733367256074, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.472391128540039, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8821477890014648, "num_tokens": 640443532.0, "step": 16789 }, { "epoch": 2.1358605775346646, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.405542373657227, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.866412878036499, "num_tokens": 640480425.0, "step": 16790 }, { "epoch": 2.135987787813255, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.481847763061523, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8753657937049866, "num_tokens": 640525658.0, "step": 16791 }, { "epoch": 2.1361149980918457, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.45184326171875, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8823394775390625, "num_tokens": 640563276.0, "step": 16792 }, { "epoch": 2.136242208370436, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.487180709838867, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.866357684135437, "num_tokens": 640608150.0, "step": 16793 }, { "epoch": 2.1363694186490267, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.493334770202637, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8748953342437744, "num_tokens": 640646127.0, "step": 16794 }, { "epoch": 2.1364966289276173, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.526945114135742, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8787425756454468, "num_tokens": 640683195.0, "step": 16795 }, { "epoch": 2.136623839206208, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.456388473510742, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8869913816452026, "num_tokens": 640724102.0, "step": 16796 }, { "epoch": 2.1367510494847983, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.469797134399414, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8692449927330017, "num_tokens": 640762624.0, "step": 16797 }, { "epoch": 2.136878259763389, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.440033912658691, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8681848645210266, "num_tokens": 640799437.0, "step": 16798 }, { "epoch": 2.1370054700419794, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.50606632232666, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8797246813774109, "num_tokens": 640837115.0, "step": 16799 }, { "epoch": 2.13713268032057, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.466358184814453, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8735182881355286, "num_tokens": 640882312.0, "step": 16800 }, { "epoch": 2.1372598905991604, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.434369087219238, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8738065361976624, "num_tokens": 640921032.0, "step": 16801 }, { "epoch": 2.137387100877751, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.539032936096191, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.874080240726471, "num_tokens": 640957424.0, "step": 16802 }, { "epoch": 2.1375143111563415, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.507110595703125, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8650941848754883, "num_tokens": 640992985.0, "step": 16803 }, { "epoch": 2.137641521434932, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.4951753616333, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8812868595123291, "num_tokens": 641030379.0, "step": 16804 }, { "epoch": 2.1377687317135226, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.532106399536133, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8653163909912109, "num_tokens": 641068526.0, "step": 16805 }, { "epoch": 2.137895941992113, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.527769088745117, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8670098185539246, "num_tokens": 641103351.0, "step": 16806 }, { "epoch": 2.1380231522707036, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.462516784667969, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8751403093338013, "num_tokens": 641143570.0, "step": 16807 }, { "epoch": 2.138150362549294, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.510866165161133, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8787753582000732, "num_tokens": 641180625.0, "step": 16808 }, { "epoch": 2.1382775728278847, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.481339454650879, "learning_rate": 1e-06, "loss": 0.473, "mean_token_accuracy": 0.8527463674545288, "num_tokens": 641226770.0, "step": 16809 }, { "epoch": 2.138404783106475, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.468853950500488, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.883939802646637, "num_tokens": 641269682.0, "step": 16810 }, { "epoch": 2.1385319933850653, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.46833610534668, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8854767084121704, "num_tokens": 641303803.0, "step": 16811 }, { "epoch": 2.1386592036636562, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.458775520324707, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8680768013000488, "num_tokens": 641343141.0, "step": 16812 }, { "epoch": 2.1387864139422463, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.548643112182617, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8769943714141846, "num_tokens": 641384006.0, "step": 16813 }, { "epoch": 2.138913624220837, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.510802268981934, "learning_rate": 1e-06, "loss": 0.4607, "mean_token_accuracy": 0.852619469165802, "num_tokens": 641423998.0, "step": 16814 }, { "epoch": 2.1390408344994274, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.462063789367676, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8719814419746399, "num_tokens": 641457886.0, "step": 16815 }, { "epoch": 2.139168044778018, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.470602035522461, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8738798499107361, "num_tokens": 641503386.0, "step": 16816 }, { "epoch": 2.1392952550566084, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.489651679992676, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8832680583000183, "num_tokens": 641539533.0, "step": 16817 }, { "epoch": 2.139422465335199, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.462532997131348, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.86806321144104, "num_tokens": 641574020.0, "step": 16818 }, { "epoch": 2.1395496756137895, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.568670272827148, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8766768574714661, "num_tokens": 641605206.0, "step": 16819 }, { "epoch": 2.13967688589238, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.490290641784668, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8632736206054688, "num_tokens": 641649857.0, "step": 16820 }, { "epoch": 2.1398040961709706, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.512886047363281, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8686686158180237, "num_tokens": 641683360.0, "step": 16821 }, { "epoch": 2.139931306449561, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.507926940917969, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8739550709724426, "num_tokens": 641720105.0, "step": 16822 }, { "epoch": 2.1400585167281516, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.471344947814941, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8757048845291138, "num_tokens": 641750718.0, "step": 16823 }, { "epoch": 2.140185727006742, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.500855445861816, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8716726303100586, "num_tokens": 641791620.0, "step": 16824 }, { "epoch": 2.1403129372853327, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.504234313964844, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.8615357875823975, "num_tokens": 641833566.0, "step": 16825 }, { "epoch": 2.140440147563923, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.464462280273438, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8799992799758911, "num_tokens": 641869080.0, "step": 16826 }, { "epoch": 2.1405673578425137, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.528115272521973, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8705713748931885, "num_tokens": 641910195.0, "step": 16827 }, { "epoch": 2.1406945681211043, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.461259841918945, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8607416749000549, "num_tokens": 641947270.0, "step": 16828 }, { "epoch": 2.140821778399695, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.518623352050781, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8738424181938171, "num_tokens": 641986472.0, "step": 16829 }, { "epoch": 2.1409489886782853, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.509400367736816, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8594372272491455, "num_tokens": 642028130.0, "step": 16830 }, { "epoch": 2.141076198956876, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.501585006713867, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8678369522094727, "num_tokens": 642067097.0, "step": 16831 }, { "epoch": 2.1412034092354664, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.53065299987793, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8843531608581543, "num_tokens": 642099628.0, "step": 16832 }, { "epoch": 2.141330619514057, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.590947151184082, "learning_rate": 1e-06, "loss": 0.3282, "mean_token_accuracy": 0.8953951001167297, "num_tokens": 642135941.0, "step": 16833 }, { "epoch": 2.1414578297926474, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.469578742980957, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8663049340248108, "num_tokens": 642169465.0, "step": 16834 }, { "epoch": 2.141585040071238, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.554659843444824, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8769499659538269, "num_tokens": 642208961.0, "step": 16835 }, { "epoch": 2.141712250349828, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.49161148071289, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8717000484466553, "num_tokens": 642249691.0, "step": 16836 }, { "epoch": 2.141839460628419, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.500319480895996, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8769286870956421, "num_tokens": 642292425.0, "step": 16837 }, { "epoch": 2.141966670907009, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.555763244628906, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8819248676300049, "num_tokens": 642330982.0, "step": 16838 }, { "epoch": 2.1420938811855996, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.520360946655273, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8751461505889893, "num_tokens": 642365414.0, "step": 16839 }, { "epoch": 2.14222109146419, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.458905220031738, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.867671549320221, "num_tokens": 642398660.0, "step": 16840 }, { "epoch": 2.1423483017427807, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.50034236907959, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8804706335067749, "num_tokens": 642432605.0, "step": 16841 }, { "epoch": 2.142475512021371, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.503331184387207, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8818407654762268, "num_tokens": 642472041.0, "step": 16842 }, { "epoch": 2.1426027222999617, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.442896842956543, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8835364580154419, "num_tokens": 642512399.0, "step": 16843 }, { "epoch": 2.1427299325785523, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.525673866271973, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8671408891677856, "num_tokens": 642550829.0, "step": 16844 }, { "epoch": 2.142857142857143, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.49710464477539, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8730310201644897, "num_tokens": 642591255.0, "step": 16845 }, { "epoch": 2.1429843531357333, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.546055793762207, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8748670816421509, "num_tokens": 642625772.0, "step": 16846 }, { "epoch": 2.143111563414324, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.407384872436523, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8671779036521912, "num_tokens": 642667596.0, "step": 16847 }, { "epoch": 2.1432387736929144, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.539867401123047, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8746554851531982, "num_tokens": 642702159.0, "step": 16848 }, { "epoch": 2.143365983971505, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.474298477172852, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8875377178192139, "num_tokens": 642737307.0, "step": 16849 }, { "epoch": 2.1434931942500954, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.47730541229248, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.863531768321991, "num_tokens": 642776660.0, "step": 16850 }, { "epoch": 2.143620404528686, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.519255638122559, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8673044443130493, "num_tokens": 642814131.0, "step": 16851 }, { "epoch": 2.1437476148072765, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.479146957397461, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.854418933391571, "num_tokens": 642854138.0, "step": 16852 }, { "epoch": 2.143874825085867, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.520028114318848, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8807245492935181, "num_tokens": 642893767.0, "step": 16853 }, { "epoch": 2.1440020353644575, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528192520141602, "learning_rate": 1e-06, "loss": 0.3473, "mean_token_accuracy": 0.8871432542800903, "num_tokens": 642928286.0, "step": 16854 }, { "epoch": 2.144129245643048, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.502494812011719, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8810315728187561, "num_tokens": 642971220.0, "step": 16855 }, { "epoch": 2.1442564559216386, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.497603416442871, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8773723244667053, "num_tokens": 643013618.0, "step": 16856 }, { "epoch": 2.144383666200229, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.515313148498535, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8766843676567078, "num_tokens": 643058122.0, "step": 16857 }, { "epoch": 2.1445108764788197, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549057960510254, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8608206510543823, "num_tokens": 643096662.0, "step": 16858 }, { "epoch": 2.14463808675741, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.431680679321289, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8855516910552979, "num_tokens": 643137923.0, "step": 16859 }, { "epoch": 2.1447652970360007, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.503034591674805, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8791942000389099, "num_tokens": 643176918.0, "step": 16860 }, { "epoch": 2.144892507314591, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.44713306427002, "learning_rate": 1e-06, "loss": 0.3478, "mean_token_accuracy": 0.8914634585380554, "num_tokens": 643217525.0, "step": 16861 }, { "epoch": 2.1450197175931813, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.460494041442871, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8899044394493103, "num_tokens": 643255131.0, "step": 16862 }, { "epoch": 2.145146927871772, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.51579761505127, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8812015056610107, "num_tokens": 643292092.0, "step": 16863 }, { "epoch": 2.1452741381503624, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.494195938110352, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8708469867706299, "num_tokens": 643330510.0, "step": 16864 }, { "epoch": 2.145401348428953, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.44433307647705, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8685575127601624, "num_tokens": 643372900.0, "step": 16865 }, { "epoch": 2.1455285587075434, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.539414405822754, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8815902471542358, "num_tokens": 643412866.0, "step": 16866 }, { "epoch": 2.145655768986134, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.440701484680176, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8573391437530518, "num_tokens": 643452846.0, "step": 16867 }, { "epoch": 2.1457829792647245, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.533514022827148, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8787375688552856, "num_tokens": 643488264.0, "step": 16868 }, { "epoch": 2.145910189543315, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.487563133239746, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.861406683921814, "num_tokens": 643528131.0, "step": 16869 }, { "epoch": 2.1460373998219056, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.535987854003906, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8737605810165405, "num_tokens": 643564041.0, "step": 16870 }, { "epoch": 2.146164610100496, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.482144355773926, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.862351655960083, "num_tokens": 643601507.0, "step": 16871 }, { "epoch": 2.1462918203790866, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.46020221710205, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8581804633140564, "num_tokens": 643647161.0, "step": 16872 }, { "epoch": 2.146419030657677, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.47240161895752, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8787649869918823, "num_tokens": 643687114.0, "step": 16873 }, { "epoch": 2.1465462409362677, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.492280960083008, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8594368696212769, "num_tokens": 643722110.0, "step": 16874 }, { "epoch": 2.146673451214858, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.557448387145996, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8511431217193604, "num_tokens": 643765770.0, "step": 16875 }, { "epoch": 2.1468006614934487, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.573241233825684, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8662772178649902, "num_tokens": 643810097.0, "step": 16876 }, { "epoch": 2.1469278717720393, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.4450101852417, "learning_rate": 1e-06, "loss": 0.4706, "mean_token_accuracy": 0.8478251695632935, "num_tokens": 643847904.0, "step": 16877 }, { "epoch": 2.14705508205063, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.564409255981445, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8812946081161499, "num_tokens": 643889157.0, "step": 16878 }, { "epoch": 2.1471822923292203, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.474913597106934, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8760564923286438, "num_tokens": 643931181.0, "step": 16879 }, { "epoch": 2.147309502607811, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528783798217773, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8715648055076599, "num_tokens": 643965415.0, "step": 16880 }, { "epoch": 2.1474367128864014, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.509573936462402, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8584938049316406, "num_tokens": 644010647.0, "step": 16881 }, { "epoch": 2.147563923164992, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.470212936401367, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8893235921859741, "num_tokens": 644044347.0, "step": 16882 }, { "epoch": 2.1476911334435824, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.523993492126465, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8803977966308594, "num_tokens": 644083456.0, "step": 16883 }, { "epoch": 2.147818343722173, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.536128997802734, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8641847372055054, "num_tokens": 644123851.0, "step": 16884 }, { "epoch": 2.1479455540007635, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.516692161560059, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.883030891418457, "num_tokens": 644168880.0, "step": 16885 }, { "epoch": 2.1480727642793536, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.575252532958984, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8753270506858826, "num_tokens": 644204532.0, "step": 16886 }, { "epoch": 2.148199974557944, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.477438926696777, "learning_rate": 1e-06, "loss": 0.3426, "mean_token_accuracy": 0.891473650932312, "num_tokens": 644246160.0, "step": 16887 }, { "epoch": 2.1483271848365346, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528611183166504, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8861328959465027, "num_tokens": 644284881.0, "step": 16888 }, { "epoch": 2.148454395115125, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.590316772460938, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8675597906112671, "num_tokens": 644326661.0, "step": 16889 }, { "epoch": 2.1485816053937157, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.54315185546875, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8778239488601685, "num_tokens": 644364420.0, "step": 16890 }, { "epoch": 2.148708815672306, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528641700744629, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.8564882278442383, "num_tokens": 644398626.0, "step": 16891 }, { "epoch": 2.1488360259508967, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.521028518676758, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8732923269271851, "num_tokens": 644434772.0, "step": 16892 }, { "epoch": 2.1489632362294873, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.535701751708984, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8710079193115234, "num_tokens": 644474425.0, "step": 16893 }, { "epoch": 2.149090446508078, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.546978950500488, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.866141676902771, "num_tokens": 644507796.0, "step": 16894 }, { "epoch": 2.1492176567866683, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549479484558105, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8694301843643188, "num_tokens": 644544172.0, "step": 16895 }, { "epoch": 2.149344867065259, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.522656440734863, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8731797337532043, "num_tokens": 644582217.0, "step": 16896 }, { "epoch": 2.1494720773438494, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.5105562210083, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8799653649330139, "num_tokens": 644620028.0, "step": 16897 }, { "epoch": 2.14959928762244, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.5585355758667, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8580482602119446, "num_tokens": 644663743.0, "step": 16898 }, { "epoch": 2.1497264979010304, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.530843734741211, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.875174343585968, "num_tokens": 644700914.0, "step": 16899 }, { "epoch": 2.149853708179621, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.506027221679688, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8700677156448364, "num_tokens": 644744232.0, "step": 16900 }, { "epoch": 2.1499809184582115, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.538063049316406, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.866949200630188, "num_tokens": 644780851.0, "step": 16901 }, { "epoch": 2.150108128736802, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.540936470031738, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8814061880111694, "num_tokens": 644822103.0, "step": 16902 }, { "epoch": 2.1502353390153925, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.480785369873047, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8703070878982544, "num_tokens": 644861209.0, "step": 16903 }, { "epoch": 2.150362549293983, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.568320274353027, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8612947463989258, "num_tokens": 644899860.0, "step": 16904 }, { "epoch": 2.1504897595725736, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.507912635803223, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8691880702972412, "num_tokens": 644943101.0, "step": 16905 }, { "epoch": 2.150616969851164, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.550849914550781, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8618078231811523, "num_tokens": 644981786.0, "step": 16906 }, { "epoch": 2.1507441801297547, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.551839828491211, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.876149594783783, "num_tokens": 645017949.0, "step": 16907 }, { "epoch": 2.150871390408345, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.542106628417969, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8719336986541748, "num_tokens": 645053443.0, "step": 16908 }, { "epoch": 2.1509986006869353, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.501011848449707, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8746778964996338, "num_tokens": 645089916.0, "step": 16909 }, { "epoch": 2.1511258109655262, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.521454811096191, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8570611476898193, "num_tokens": 645134531.0, "step": 16910 }, { "epoch": 2.1512530212441163, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.527369499206543, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8683984279632568, "num_tokens": 645172608.0, "step": 16911 }, { "epoch": 2.151380231522707, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.501138687133789, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8689290285110474, "num_tokens": 645209895.0, "step": 16912 }, { "epoch": 2.1515074418012974, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528386116027832, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8793364763259888, "num_tokens": 645245259.0, "step": 16913 }, { "epoch": 2.151634652079888, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.476975440979004, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8676745891571045, "num_tokens": 645284962.0, "step": 16914 }, { "epoch": 2.1517618623584784, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.494658470153809, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8879450559616089, "num_tokens": 645322352.0, "step": 16915 }, { "epoch": 2.151889072637069, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.4307861328125, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.871886670589447, "num_tokens": 645358401.0, "step": 16916 }, { "epoch": 2.1520162829156595, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.53750991821289, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.885273277759552, "num_tokens": 645399147.0, "step": 16917 }, { "epoch": 2.15214349319425, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.487919807434082, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8548194766044617, "num_tokens": 645437286.0, "step": 16918 }, { "epoch": 2.1522707034728406, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.52053165435791, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8660817742347717, "num_tokens": 645477853.0, "step": 16919 }, { "epoch": 2.152397913751431, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.562832832336426, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8628242015838623, "num_tokens": 645520476.0, "step": 16920 }, { "epoch": 2.1525251240300216, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.523656845092773, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.859129011631012, "num_tokens": 645561539.0, "step": 16921 }, { "epoch": 2.152652334308612, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.49179744720459, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8684621453285217, "num_tokens": 645602950.0, "step": 16922 }, { "epoch": 2.1527795445872027, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.56593132019043, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8630231022834778, "num_tokens": 645639215.0, "step": 16923 }, { "epoch": 2.152906754865793, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.512866973876953, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8717455267906189, "num_tokens": 645677287.0, "step": 16924 }, { "epoch": 2.1530339651443837, "ewc_loss": 0.034912109375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.505220413208008, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8797495365142822, "num_tokens": 645710784.0, "step": 16925 }, { "epoch": 2.1531611754229742, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.62419605255127, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8684573769569397, "num_tokens": 645750568.0, "step": 16926 }, { "epoch": 2.1532883857015648, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.537524223327637, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8779064416885376, "num_tokens": 645783911.0, "step": 16927 }, { "epoch": 2.1534155959801553, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.536774635314941, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.888577401638031, "num_tokens": 645822524.0, "step": 16928 }, { "epoch": 2.153542806258746, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.609477996826172, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8684710264205933, "num_tokens": 645856924.0, "step": 16929 }, { "epoch": 2.1536700165373364, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.570319175720215, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8658560514450073, "num_tokens": 645896995.0, "step": 16930 }, { "epoch": 2.153797226815927, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.487070083618164, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8792234659194946, "num_tokens": 645941768.0, "step": 16931 }, { "epoch": 2.1539244370945174, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.572972297668457, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8713721036911011, "num_tokens": 645981909.0, "step": 16932 }, { "epoch": 2.154051647373108, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.515252113342285, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8637677431106567, "num_tokens": 646022034.0, "step": 16933 }, { "epoch": 2.154178857651698, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.652843475341797, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8717787861824036, "num_tokens": 646056714.0, "step": 16934 }, { "epoch": 2.1543060679302886, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.466670036315918, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8643743395805359, "num_tokens": 646095668.0, "step": 16935 }, { "epoch": 2.154433278208879, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.613665580749512, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8567315340042114, "num_tokens": 646136745.0, "step": 16936 }, { "epoch": 2.1545604884874696, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.515603065490723, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.869533896446228, "num_tokens": 646180314.0, "step": 16937 }, { "epoch": 2.15468769876606, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.545536041259766, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.874665379524231, "num_tokens": 646218191.0, "step": 16938 }, { "epoch": 2.1548149090446507, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.475756645202637, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8866212368011475, "num_tokens": 646256350.0, "step": 16939 }, { "epoch": 2.154942119323241, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.504313468933105, "learning_rate": 1e-06, "loss": 0.3222, "mean_token_accuracy": 0.8949068188667297, "num_tokens": 646290915.0, "step": 16940 }, { "epoch": 2.1550693296018317, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.606178283691406, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8658767342567444, "num_tokens": 646325838.0, "step": 16941 }, { "epoch": 2.1551965398804223, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528010368347168, "learning_rate": 1e-06, "loss": 0.4591, "mean_token_accuracy": 0.8548609018325806, "num_tokens": 646362742.0, "step": 16942 }, { "epoch": 2.155323750159013, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.545583724975586, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8596181273460388, "num_tokens": 646403958.0, "step": 16943 }, { "epoch": 2.1554509604376033, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.561392784118652, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8663601875305176, "num_tokens": 646442872.0, "step": 16944 }, { "epoch": 2.155578170716194, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.586698532104492, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8791362047195435, "num_tokens": 646484642.0, "step": 16945 }, { "epoch": 2.1557053809947844, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.502314567565918, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8814952373504639, "num_tokens": 646521688.0, "step": 16946 }, { "epoch": 2.155832591273375, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.471414566040039, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8796348571777344, "num_tokens": 646560182.0, "step": 16947 }, { "epoch": 2.1559598015519654, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.558425903320312, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.87698894739151, "num_tokens": 646593252.0, "step": 16948 }, { "epoch": 2.156087011830556, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.503253936767578, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8699109554290771, "num_tokens": 646635514.0, "step": 16949 }, { "epoch": 2.1562142221091465, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.544487953186035, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8597928285598755, "num_tokens": 646674785.0, "step": 16950 }, { "epoch": 2.156341432387737, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.472949981689453, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8728954195976257, "num_tokens": 646713832.0, "step": 16951 }, { "epoch": 2.1564686426663275, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.480703353881836, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8758513927459717, "num_tokens": 646752218.0, "step": 16952 }, { "epoch": 2.156595852944918, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.463449478149414, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8715196847915649, "num_tokens": 646786196.0, "step": 16953 }, { "epoch": 2.1567230632235086, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549026489257812, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.865891695022583, "num_tokens": 646821462.0, "step": 16954 }, { "epoch": 2.156850273502099, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.505810737609863, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8799124956130981, "num_tokens": 646856613.0, "step": 16955 }, { "epoch": 2.1569774837806897, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.488763809204102, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8833556175231934, "num_tokens": 646893013.0, "step": 16956 }, { "epoch": 2.15710469405928, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.4822416305542, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8775784969329834, "num_tokens": 646936180.0, "step": 16957 }, { "epoch": 2.1572319043378707, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.553871154785156, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8725897669792175, "num_tokens": 646979314.0, "step": 16958 }, { "epoch": 2.157359114616461, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.451043128967285, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8819187879562378, "num_tokens": 647020136.0, "step": 16959 }, { "epoch": 2.1574863248950513, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.54257869720459, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8769433498382568, "num_tokens": 647052970.0, "step": 16960 }, { "epoch": 2.157613535173642, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.490562438964844, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8667517900466919, "num_tokens": 647093195.0, "step": 16961 }, { "epoch": 2.1577407454522324, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.500139236450195, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8727976083755493, "num_tokens": 647133800.0, "step": 16962 }, { "epoch": 2.157867955730823, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.521147727966309, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8563849925994873, "num_tokens": 647167704.0, "step": 16963 }, { "epoch": 2.1579951660094134, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.559674263000488, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8598213195800781, "num_tokens": 647208474.0, "step": 16964 }, { "epoch": 2.158122376288004, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.473291397094727, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8860538005828857, "num_tokens": 647242131.0, "step": 16965 }, { "epoch": 2.1582495865665945, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.555724143981934, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8545110821723938, "num_tokens": 647280952.0, "step": 16966 }, { "epoch": 2.158376796845185, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.4574556350708, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8679839372634888, "num_tokens": 647318411.0, "step": 16967 }, { "epoch": 2.1585040071237755, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.528084754943848, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8600959181785583, "num_tokens": 647361002.0, "step": 16968 }, { "epoch": 2.158631217402366, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.537864685058594, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8762074708938599, "num_tokens": 647402995.0, "step": 16969 }, { "epoch": 2.1587584276809566, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.530226707458496, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8746840953826904, "num_tokens": 647441220.0, "step": 16970 }, { "epoch": 2.158885637959547, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.55379581451416, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8653878569602966, "num_tokens": 647476832.0, "step": 16971 }, { "epoch": 2.1590128482381377, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.491758346557617, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.874754786491394, "num_tokens": 647516855.0, "step": 16972 }, { "epoch": 2.159140058516728, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.502755165100098, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8694618940353394, "num_tokens": 647556913.0, "step": 16973 }, { "epoch": 2.1592672687953187, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.539618492126465, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8888164162635803, "num_tokens": 647591242.0, "step": 16974 }, { "epoch": 2.1593944790739092, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.537620544433594, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8760823607444763, "num_tokens": 647632156.0, "step": 16975 }, { "epoch": 2.1595216893524998, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.538368225097656, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8785079121589661, "num_tokens": 647669857.0, "step": 16976 }, { "epoch": 2.1596488996310903, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.537310600280762, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8599752187728882, "num_tokens": 647715580.0, "step": 16977 }, { "epoch": 2.159776109909681, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.56559944152832, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8636904358863831, "num_tokens": 647755917.0, "step": 16978 }, { "epoch": 2.1599033201882714, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.555418014526367, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8887624144554138, "num_tokens": 647802205.0, "step": 16979 }, { "epoch": 2.160030530466862, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.48908519744873, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8588989973068237, "num_tokens": 647846627.0, "step": 16980 }, { "epoch": 2.1601577407454524, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.557819366455078, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8638349771499634, "num_tokens": 647885033.0, "step": 16981 }, { "epoch": 2.160284951024043, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.599360466003418, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.862733006477356, "num_tokens": 647923370.0, "step": 16982 }, { "epoch": 2.1604121613026335, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.5053071975708, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8783934712409973, "num_tokens": 647963431.0, "step": 16983 }, { "epoch": 2.1605393715812236, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.559863090515137, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8641442060470581, "num_tokens": 648002854.0, "step": 16984 }, { "epoch": 2.160666581859814, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.610917091369629, "learning_rate": 1e-06, "loss": 0.4477, "mean_token_accuracy": 0.8613316416740417, "num_tokens": 648041082.0, "step": 16985 }, { "epoch": 2.1607937921384046, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.548784255981445, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8644427061080933, "num_tokens": 648082109.0, "step": 16986 }, { "epoch": 2.160921002416995, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.574488639831543, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8864233493804932, "num_tokens": 648117398.0, "step": 16987 }, { "epoch": 2.1610482126955857, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.548382759094238, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8797212243080139, "num_tokens": 648158997.0, "step": 16988 }, { "epoch": 2.161175422974176, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.524145126342773, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.870194137096405, "num_tokens": 648194334.0, "step": 16989 }, { "epoch": 2.1613026332527667, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.628300666809082, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8690598607063293, "num_tokens": 648230601.0, "step": 16990 }, { "epoch": 2.1614298435313573, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.51087760925293, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.865912675857544, "num_tokens": 648274979.0, "step": 16991 }, { "epoch": 2.161557053809948, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.471856117248535, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8569389581680298, "num_tokens": 648319229.0, "step": 16992 }, { "epoch": 2.1616842640885383, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.570530891418457, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8841866254806519, "num_tokens": 648354550.0, "step": 16993 }, { "epoch": 2.161811474367129, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.502382278442383, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8848503828048706, "num_tokens": 648389820.0, "step": 16994 }, { "epoch": 2.1619386846457194, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.572835922241211, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8862660527229309, "num_tokens": 648423841.0, "step": 16995 }, { "epoch": 2.16206589492431, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.56698989868164, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8606365919113159, "num_tokens": 648459865.0, "step": 16996 }, { "epoch": 2.1621931052029004, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.44534683227539, "learning_rate": 1e-06, "loss": 0.339, "mean_token_accuracy": 0.8894228935241699, "num_tokens": 648492639.0, "step": 16997 }, { "epoch": 2.162320315481491, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.571131706237793, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8772792816162109, "num_tokens": 648533456.0, "step": 16998 }, { "epoch": 2.1624475257600815, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549346923828125, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8854494094848633, "num_tokens": 648571955.0, "step": 16999 }, { "epoch": 2.162574736038672, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549881935119629, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.8890928626060486, "num_tokens": 648614284.0, "step": 17000 }, { "epoch": 2.1627019463172625, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.541312217712402, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8806018829345703, "num_tokens": 648654399.0, "step": 17001 }, { "epoch": 2.162829156595853, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.56643009185791, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8750197291374207, "num_tokens": 648694490.0, "step": 17002 }, { "epoch": 2.1629563668744436, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.56083869934082, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8779680728912354, "num_tokens": 648730665.0, "step": 17003 }, { "epoch": 2.163083577153034, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 13.472959518432617, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8770372867584229, "num_tokens": 648765289.0, "step": 17004 }, { "epoch": 2.1632107874316246, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.584285736083984, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8734137415885925, "num_tokens": 648803982.0, "step": 17005 }, { "epoch": 2.163337997710215, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.537945747375488, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8774493932723999, "num_tokens": 648840992.0, "step": 17006 }, { "epoch": 2.1634652079888053, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.556415557861328, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8681896328926086, "num_tokens": 648883621.0, "step": 17007 }, { "epoch": 2.1635924182673962, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.607246398925781, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8581478595733643, "num_tokens": 648922258.0, "step": 17008 }, { "epoch": 2.1637196285459863, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.506505012512207, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8724057674407959, "num_tokens": 648965010.0, "step": 17009 }, { "epoch": 2.163846838824577, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.561914443969727, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8681308031082153, "num_tokens": 649008124.0, "step": 17010 }, { "epoch": 2.1639740491031674, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.528555870056152, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8723087310791016, "num_tokens": 649044743.0, "step": 17011 }, { "epoch": 2.164101259381758, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.539525032043457, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8853086233139038, "num_tokens": 649078410.0, "step": 17012 }, { "epoch": 2.1642284696603484, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.525349617004395, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8837124109268188, "num_tokens": 649115415.0, "step": 17013 }, { "epoch": 2.164355679938939, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.505417823791504, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8705548048019409, "num_tokens": 649156266.0, "step": 17014 }, { "epoch": 2.1644828902175295, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.546449661254883, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.879234790802002, "num_tokens": 649192552.0, "step": 17015 }, { "epoch": 2.16461010049612, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.507037162780762, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8753364682197571, "num_tokens": 649224163.0, "step": 17016 }, { "epoch": 2.1647373107747105, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.55569076538086, "learning_rate": 1e-06, "loss": 0.3335, "mean_token_accuracy": 0.8920164704322815, "num_tokens": 649259348.0, "step": 17017 }, { "epoch": 2.164864521053301, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.497922897338867, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.887673556804657, "num_tokens": 649296875.0, "step": 17018 }, { "epoch": 2.1649917313318916, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.610873222351074, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8616122603416443, "num_tokens": 649333455.0, "step": 17019 }, { "epoch": 2.165118941610482, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.53930950164795, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8740823268890381, "num_tokens": 649364567.0, "step": 17020 }, { "epoch": 2.1652461518890727, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.571741104125977, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8553805351257324, "num_tokens": 649400579.0, "step": 17021 }, { "epoch": 2.165373362167663, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.543229103088379, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8853229880332947, "num_tokens": 649432940.0, "step": 17022 }, { "epoch": 2.1655005724462537, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.558367729187012, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8768138885498047, "num_tokens": 649466695.0, "step": 17023 }, { "epoch": 2.1656277827248442, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.55273723602295, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8713634014129639, "num_tokens": 649504578.0, "step": 17024 }, { "epoch": 2.1657549930034348, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.581625938415527, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8625125885009766, "num_tokens": 649541536.0, "step": 17025 }, { "epoch": 2.1658822032820253, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.510343551635742, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8708560466766357, "num_tokens": 649581280.0, "step": 17026 }, { "epoch": 2.166009413560616, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.576285362243652, "learning_rate": 1e-06, "loss": 0.3395, "mean_token_accuracy": 0.8944646120071411, "num_tokens": 649619837.0, "step": 17027 }, { "epoch": 2.1661366238392064, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.55636978149414, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8847854137420654, "num_tokens": 649654708.0, "step": 17028 }, { "epoch": 2.166263834117797, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.620199203491211, "learning_rate": 1e-06, "loss": 0.4564, "mean_token_accuracy": 0.8526651263237, "num_tokens": 649691709.0, "step": 17029 }, { "epoch": 2.1663910443963874, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.544447898864746, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8695365190505981, "num_tokens": 649734876.0, "step": 17030 }, { "epoch": 2.166518254674978, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.575252532958984, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8816044330596924, "num_tokens": 649775477.0, "step": 17031 }, { "epoch": 2.166645464953568, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.555798530578613, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8853335380554199, "num_tokens": 649816136.0, "step": 17032 }, { "epoch": 2.1667726752321586, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.610663414001465, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8823355436325073, "num_tokens": 649847942.0, "step": 17033 }, { "epoch": 2.166899885510749, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.56397533416748, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.869366466999054, "num_tokens": 649887875.0, "step": 17034 }, { "epoch": 2.1670270957893396, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.623136520385742, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8603857755661011, "num_tokens": 649920557.0, "step": 17035 }, { "epoch": 2.16715430606793, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.565286636352539, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8814981579780579, "num_tokens": 649959704.0, "step": 17036 }, { "epoch": 2.1672815163465207, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.593976020812988, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.878503680229187, "num_tokens": 650000433.0, "step": 17037 }, { "epoch": 2.167408726625111, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.5538330078125, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8854707479476929, "num_tokens": 650038793.0, "step": 17038 }, { "epoch": 2.1675359369037017, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.622490882873535, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8818625807762146, "num_tokens": 650072877.0, "step": 17039 }, { "epoch": 2.1676631471822922, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.56485366821289, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8614584803581238, "num_tokens": 650109245.0, "step": 17040 }, { "epoch": 2.1677903574608828, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.54345703125, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8865574598312378, "num_tokens": 650150254.0, "step": 17041 }, { "epoch": 2.1679175677394733, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.647518157958984, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8551006317138672, "num_tokens": 650189587.0, "step": 17042 }, { "epoch": 2.168044778018064, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.584454536437988, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8696572780609131, "num_tokens": 650225443.0, "step": 17043 }, { "epoch": 2.1681719882966544, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.557673454284668, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8692956566810608, "num_tokens": 650262319.0, "step": 17044 }, { "epoch": 2.168299198575245, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.596334457397461, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8582395911216736, "num_tokens": 650297137.0, "step": 17045 }, { "epoch": 2.1684264088538354, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.550610542297363, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8726903200149536, "num_tokens": 650335521.0, "step": 17046 }, { "epoch": 2.168553619132426, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.519149780273438, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8679901957511902, "num_tokens": 650373050.0, "step": 17047 }, { "epoch": 2.1686808294110165, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.574610710144043, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8793816566467285, "num_tokens": 650407438.0, "step": 17048 }, { "epoch": 2.168808039689607, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.599871635437012, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8680968284606934, "num_tokens": 650447793.0, "step": 17049 }, { "epoch": 2.1689352499681975, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.582589149475098, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8693414330482483, "num_tokens": 650487167.0, "step": 17050 }, { "epoch": 2.169062460246788, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.519339561462402, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8609454035758972, "num_tokens": 650522960.0, "step": 17051 }, { "epoch": 2.1691896705253786, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.564628601074219, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8672381639480591, "num_tokens": 650564161.0, "step": 17052 }, { "epoch": 2.169316880803969, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.56222915649414, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.879637598991394, "num_tokens": 650599201.0, "step": 17053 }, { "epoch": 2.1694440910825596, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.568414688110352, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8540834784507751, "num_tokens": 650637385.0, "step": 17054 }, { "epoch": 2.16957130136115, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.586526870727539, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8736037015914917, "num_tokens": 650671954.0, "step": 17055 }, { "epoch": 2.1696985116397407, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.606123924255371, "learning_rate": 1e-06, "loss": 0.3443, "mean_token_accuracy": 0.8902102708816528, "num_tokens": 650708475.0, "step": 17056 }, { "epoch": 2.169825721918331, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.562896728515625, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8766785860061646, "num_tokens": 650748089.0, "step": 17057 }, { "epoch": 2.1699529321969213, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.527565002441406, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8750819563865662, "num_tokens": 650790869.0, "step": 17058 }, { "epoch": 2.170080142475512, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.564329147338867, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8601139783859253, "num_tokens": 650830697.0, "step": 17059 }, { "epoch": 2.1702073527541024, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549273490905762, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8820737600326538, "num_tokens": 650868755.0, "step": 17060 }, { "epoch": 2.170334563032693, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.53909683227539, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8833532333374023, "num_tokens": 650908019.0, "step": 17061 }, { "epoch": 2.1704617733112834, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.566116333007812, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.872606635093689, "num_tokens": 650954195.0, "step": 17062 }, { "epoch": 2.170588983589874, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.572681427001953, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8859856128692627, "num_tokens": 650992740.0, "step": 17063 }, { "epoch": 2.1707161938684645, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.583873748779297, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8865183591842651, "num_tokens": 651031524.0, "step": 17064 }, { "epoch": 2.170843404147055, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.625329971313477, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8711337447166443, "num_tokens": 651070004.0, "step": 17065 }, { "epoch": 2.1709706144256455, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.62557315826416, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8598051071166992, "num_tokens": 651104474.0, "step": 17066 }, { "epoch": 2.171097824704236, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.593632698059082, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8666244745254517, "num_tokens": 651148913.0, "step": 17067 }, { "epoch": 2.1712250349828266, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.50793743133545, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8814204335212708, "num_tokens": 651192072.0, "step": 17068 }, { "epoch": 2.171352245261417, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.675566673278809, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8648730516433716, "num_tokens": 651226132.0, "step": 17069 }, { "epoch": 2.1714794555400077, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549860954284668, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8760166168212891, "num_tokens": 651263467.0, "step": 17070 }, { "epoch": 2.171606665818598, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.537848472595215, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8650684952735901, "num_tokens": 651309521.0, "step": 17071 }, { "epoch": 2.1717338760971887, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.582544326782227, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8762484788894653, "num_tokens": 651347026.0, "step": 17072 }, { "epoch": 2.1718610863757792, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.6232328414917, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8821630477905273, "num_tokens": 651380545.0, "step": 17073 }, { "epoch": 2.1719882966543698, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.608285903930664, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8777065277099609, "num_tokens": 651409428.0, "step": 17074 }, { "epoch": 2.1721155069329603, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.608920097351074, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8605632185935974, "num_tokens": 651447993.0, "step": 17075 }, { "epoch": 2.172242717211551, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.644771575927734, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8677310943603516, "num_tokens": 651488210.0, "step": 17076 }, { "epoch": 2.1723699274901414, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.563714027404785, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8692657947540283, "num_tokens": 651525742.0, "step": 17077 }, { "epoch": 2.172497137768732, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.63974666595459, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8800808191299438, "num_tokens": 651565569.0, "step": 17078 }, { "epoch": 2.1726243480473224, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.572301864624023, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8688526153564453, "num_tokens": 651605291.0, "step": 17079 }, { "epoch": 2.172751558325913, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.58837890625, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8738051056861877, "num_tokens": 651641998.0, "step": 17080 }, { "epoch": 2.1728787686045035, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.592923164367676, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8701468706130981, "num_tokens": 651683040.0, "step": 17081 }, { "epoch": 2.1730059788830935, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.629100799560547, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8816958665847778, "num_tokens": 651718067.0, "step": 17082 }, { "epoch": 2.173133189161684, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.533949851989746, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8877884745597839, "num_tokens": 651761995.0, "step": 17083 }, { "epoch": 2.1732603994402746, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.660850524902344, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8865486979484558, "num_tokens": 651797780.0, "step": 17084 }, { "epoch": 2.173387609718865, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.536221504211426, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.862835705280304, "num_tokens": 651832931.0, "step": 17085 }, { "epoch": 2.1735148199974557, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.632455825805664, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8673322200775146, "num_tokens": 651868440.0, "step": 17086 }, { "epoch": 2.173642030276046, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.622978210449219, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.876873254776001, "num_tokens": 651900029.0, "step": 17087 }, { "epoch": 2.1737692405546367, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.557036399841309, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8698426485061646, "num_tokens": 651944892.0, "step": 17088 }, { "epoch": 2.1738964508332272, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.571931838989258, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8807288408279419, "num_tokens": 651979330.0, "step": 17089 }, { "epoch": 2.1740236611118178, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.555624961853027, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8653836250305176, "num_tokens": 652017389.0, "step": 17090 }, { "epoch": 2.1741508713904083, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.594280242919922, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8692627549171448, "num_tokens": 652052260.0, "step": 17091 }, { "epoch": 2.174278081668999, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.583561897277832, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8583452701568604, "num_tokens": 652089917.0, "step": 17092 }, { "epoch": 2.1744052919475894, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.534947395324707, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.875762939453125, "num_tokens": 652129849.0, "step": 17093 }, { "epoch": 2.17453250222618, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.602994918823242, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8658560514450073, "num_tokens": 652163178.0, "step": 17094 }, { "epoch": 2.1746597125047704, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.53627872467041, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8784865140914917, "num_tokens": 652201521.0, "step": 17095 }, { "epoch": 2.174786922783361, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.548876762390137, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8740382194519043, "num_tokens": 652234820.0, "step": 17096 }, { "epoch": 2.1749141330619515, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.541708946228027, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8698002099990845, "num_tokens": 652275686.0, "step": 17097 }, { "epoch": 2.175041343340542, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.549432754516602, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.864722490310669, "num_tokens": 652313243.0, "step": 17098 }, { "epoch": 2.1751685536191325, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.556227684020996, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8877863883972168, "num_tokens": 652351550.0, "step": 17099 }, { "epoch": 2.175295763897723, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.591936111450195, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8701512813568115, "num_tokens": 652399834.0, "step": 17100 }, { "epoch": 2.1754229741763136, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.583155632019043, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8595781326293945, "num_tokens": 652435839.0, "step": 17101 }, { "epoch": 2.175550184454904, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.524852752685547, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.869477391242981, "num_tokens": 652475747.0, "step": 17102 }, { "epoch": 2.1756773947334946, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.599925994873047, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8846210241317749, "num_tokens": 652517147.0, "step": 17103 }, { "epoch": 2.175804605012085, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.582648277282715, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8854631185531616, "num_tokens": 652552811.0, "step": 17104 }, { "epoch": 2.1759318152906753, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.5842866897583, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8805944323539734, "num_tokens": 652587709.0, "step": 17105 }, { "epoch": 2.1760590255692662, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.56466293334961, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8725127577781677, "num_tokens": 652626213.0, "step": 17106 }, { "epoch": 2.1761862358478563, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.594329833984375, "learning_rate": 1e-06, "loss": 0.4949, "mean_token_accuracy": 0.8443596363067627, "num_tokens": 652665692.0, "step": 17107 }, { "epoch": 2.176313446126447, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.578174591064453, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8648458123207092, "num_tokens": 652705515.0, "step": 17108 }, { "epoch": 2.1764406564050374, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.475829124450684, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8848005533218384, "num_tokens": 652743343.0, "step": 17109 }, { "epoch": 2.176567866683628, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.58169937133789, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8678903579711914, "num_tokens": 652784587.0, "step": 17110 }, { "epoch": 2.1766950769622184, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.510258674621582, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8645740151405334, "num_tokens": 652821985.0, "step": 17111 }, { "epoch": 2.176822287240809, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.61867904663086, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8828670978546143, "num_tokens": 652855772.0, "step": 17112 }, { "epoch": 2.1769494975193995, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.55783462524414, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8688426613807678, "num_tokens": 652890497.0, "step": 17113 }, { "epoch": 2.17707670779799, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.54660415649414, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8830937743186951, "num_tokens": 652931611.0, "step": 17114 }, { "epoch": 2.1772039180765805, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.590009689331055, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8732784986495972, "num_tokens": 652972129.0, "step": 17115 }, { "epoch": 2.177331128355171, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.61137866973877, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8671193718910217, "num_tokens": 653009848.0, "step": 17116 }, { "epoch": 2.1774583386337616, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.593083381652832, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8692971467971802, "num_tokens": 653044614.0, "step": 17117 }, { "epoch": 2.177585548912352, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.590527534484863, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8861987590789795, "num_tokens": 653087614.0, "step": 17118 }, { "epoch": 2.1777127591909426, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.578158378601074, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8857812881469727, "num_tokens": 653120434.0, "step": 17119 }, { "epoch": 2.177839969469533, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.618971824645996, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8832489848136902, "num_tokens": 653159390.0, "step": 17120 }, { "epoch": 2.1779671797481237, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.566962242126465, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8657881617546082, "num_tokens": 653201045.0, "step": 17121 }, { "epoch": 2.1780943900267142, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.55081558227539, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8757622241973877, "num_tokens": 653244968.0, "step": 17122 }, { "epoch": 2.1782216003053048, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.612236976623535, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8663997650146484, "num_tokens": 653281157.0, "step": 17123 }, { "epoch": 2.1783488105838953, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.590977668762207, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.886056661605835, "num_tokens": 653321055.0, "step": 17124 }, { "epoch": 2.178476020862486, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.559992790222168, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.859837532043457, "num_tokens": 653359320.0, "step": 17125 }, { "epoch": 2.1786032311410763, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.615896224975586, "learning_rate": 1e-06, "loss": 0.3339, "mean_token_accuracy": 0.8922698497772217, "num_tokens": 653404186.0, "step": 17126 }, { "epoch": 2.178730441419667, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.545896530151367, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8832238912582397, "num_tokens": 653438122.0, "step": 17127 }, { "epoch": 2.1788576516982574, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.613765716552734, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8729806542396545, "num_tokens": 653474727.0, "step": 17128 }, { "epoch": 2.178984861976848, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.58735466003418, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8786226511001587, "num_tokens": 653513531.0, "step": 17129 }, { "epoch": 2.179112072255438, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.600909233093262, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8841250538825989, "num_tokens": 653555871.0, "step": 17130 }, { "epoch": 2.1792392825340285, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.581902503967285, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8745778799057007, "num_tokens": 653598825.0, "step": 17131 }, { "epoch": 2.179366492812619, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.669454574584961, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8824625015258789, "num_tokens": 653639838.0, "step": 17132 }, { "epoch": 2.1794937030912096, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.540511131286621, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8651593327522278, "num_tokens": 653678958.0, "step": 17133 }, { "epoch": 2.1796209133698, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.562073707580566, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8812501430511475, "num_tokens": 653718043.0, "step": 17134 }, { "epoch": 2.1797481236483907, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.59311294555664, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.8882980942726135, "num_tokens": 653756308.0, "step": 17135 }, { "epoch": 2.179875333926981, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.585627555847168, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8824416399002075, "num_tokens": 653793140.0, "step": 17136 }, { "epoch": 2.1800025442055717, "ewc_loss": 0.03515625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.60858154296875, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8793615698814392, "num_tokens": 653823441.0, "step": 17137 }, { "epoch": 2.1801297544841622, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.630304336547852, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.864590585231781, "num_tokens": 653860743.0, "step": 17138 }, { "epoch": 2.1802569647627528, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.556425094604492, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8857779502868652, "num_tokens": 653894785.0, "step": 17139 }, { "epoch": 2.1803841750413433, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.583447456359863, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8707501292228699, "num_tokens": 653932208.0, "step": 17140 }, { "epoch": 2.180511385319934, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.609440803527832, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.8600399494171143, "num_tokens": 653968818.0, "step": 17141 }, { "epoch": 2.1806385955985244, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.576732635498047, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8703513145446777, "num_tokens": 654010696.0, "step": 17142 }, { "epoch": 2.180765805877115, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.556827545166016, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.878605842590332, "num_tokens": 654047632.0, "step": 17143 }, { "epoch": 2.1808930161557054, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.626728057861328, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8775016665458679, "num_tokens": 654088200.0, "step": 17144 }, { "epoch": 2.181020226434296, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.546789169311523, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.870621383190155, "num_tokens": 654124225.0, "step": 17145 }, { "epoch": 2.1811474367128865, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.535861015319824, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8762685060501099, "num_tokens": 654163317.0, "step": 17146 }, { "epoch": 2.181274646991477, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.541644096374512, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.866131603717804, "num_tokens": 654199208.0, "step": 17147 }, { "epoch": 2.1814018572700675, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.592511177062988, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8623644709587097, "num_tokens": 654239988.0, "step": 17148 }, { "epoch": 2.181529067548658, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.530975341796875, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.86833256483078, "num_tokens": 654276361.0, "step": 17149 }, { "epoch": 2.1816562778272486, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.59361457824707, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.877143144607544, "num_tokens": 654316690.0, "step": 17150 }, { "epoch": 2.181783488105839, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.529187202453613, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.882118821144104, "num_tokens": 654354112.0, "step": 17151 }, { "epoch": 2.1819106983844296, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.56240463256836, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8848291635513306, "num_tokens": 654398896.0, "step": 17152 }, { "epoch": 2.18203790866302, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.546367645263672, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8855762481689453, "num_tokens": 654435180.0, "step": 17153 }, { "epoch": 2.1821651189416107, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.594834327697754, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8627209067344666, "num_tokens": 654473958.0, "step": 17154 }, { "epoch": 2.1822923292202008, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.584115028381348, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8602889180183411, "num_tokens": 654512918.0, "step": 17155 }, { "epoch": 2.1824195394987913, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.583152770996094, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8813214898109436, "num_tokens": 654552296.0, "step": 17156 }, { "epoch": 2.182546749777382, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.533047676086426, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.880605936050415, "num_tokens": 654592088.0, "step": 17157 }, { "epoch": 2.1826739600559724, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.601972579956055, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.866613507270813, "num_tokens": 654635119.0, "step": 17158 }, { "epoch": 2.182801170334563, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.496905326843262, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8625831604003906, "num_tokens": 654671204.0, "step": 17159 }, { "epoch": 2.1829283806131534, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.66716480255127, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.872234582901001, "num_tokens": 654702986.0, "step": 17160 }, { "epoch": 2.183055590891744, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.549890518188477, "learning_rate": 1e-06, "loss": 0.4697, "mean_token_accuracy": 0.850663423538208, "num_tokens": 654745813.0, "step": 17161 }, { "epoch": 2.1831828011703345, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.5866117477417, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8738160729408264, "num_tokens": 654783305.0, "step": 17162 }, { "epoch": 2.183310011448925, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.599496841430664, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8816578388214111, "num_tokens": 654823427.0, "step": 17163 }, { "epoch": 2.1834372217275155, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.576994895935059, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.886135458946228, "num_tokens": 654862400.0, "step": 17164 }, { "epoch": 2.183564432006106, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.525883674621582, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8783149719238281, "num_tokens": 654902484.0, "step": 17165 }, { "epoch": 2.1836916422846966, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.673725128173828, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8801061511039734, "num_tokens": 654946882.0, "step": 17166 }, { "epoch": 2.183818852563287, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.530991554260254, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8643691539764404, "num_tokens": 654977850.0, "step": 17167 }, { "epoch": 2.1839460628418776, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.643143653869629, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8596810102462769, "num_tokens": 655015352.0, "step": 17168 }, { "epoch": 2.184073273120468, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.599713325500488, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8741533756256104, "num_tokens": 655047488.0, "step": 17169 }, { "epoch": 2.1842004833990587, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.601431846618652, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8714475631713867, "num_tokens": 655083410.0, "step": 17170 }, { "epoch": 2.1843276936776492, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.64307689666748, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8816465735435486, "num_tokens": 655120594.0, "step": 17171 }, { "epoch": 2.1844549039562398, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.622527122497559, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8719055652618408, "num_tokens": 655157310.0, "step": 17172 }, { "epoch": 2.1845821142348303, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.573075294494629, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8829748034477234, "num_tokens": 655193749.0, "step": 17173 }, { "epoch": 2.184709324513421, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.646810531616211, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8880590200424194, "num_tokens": 655235532.0, "step": 17174 }, { "epoch": 2.1848365347920113, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.5598783493042, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8794021010398865, "num_tokens": 655282154.0, "step": 17175 }, { "epoch": 2.184963745070602, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.59913444519043, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8676705360412598, "num_tokens": 655321342.0, "step": 17176 }, { "epoch": 2.1850909553491924, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.627573013305664, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8906729221343994, "num_tokens": 655358107.0, "step": 17177 }, { "epoch": 2.185218165627783, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.6568021774292, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8683731555938721, "num_tokens": 655393462.0, "step": 17178 }, { "epoch": 2.1853453759063735, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.588229179382324, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8760133385658264, "num_tokens": 655436114.0, "step": 17179 }, { "epoch": 2.1854725861849635, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.66148853302002, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8759679198265076, "num_tokens": 655473119.0, "step": 17180 }, { "epoch": 2.185599796463554, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.577163696289062, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8687235116958618, "num_tokens": 655515537.0, "step": 17181 }, { "epoch": 2.1857270067421446, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.666629791259766, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8676552772521973, "num_tokens": 655555944.0, "step": 17182 }, { "epoch": 2.185854217020735, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.625306129455566, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8578547835350037, "num_tokens": 655589952.0, "step": 17183 }, { "epoch": 2.1859814272993257, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.622319221496582, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8870063424110413, "num_tokens": 655628269.0, "step": 17184 }, { "epoch": 2.186108637577916, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.600879669189453, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8763711452484131, "num_tokens": 655657315.0, "step": 17185 }, { "epoch": 2.1862358478565067, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.602005004882812, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8733899593353271, "num_tokens": 655696117.0, "step": 17186 }, { "epoch": 2.1863630581350972, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.545443534851074, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8571833372116089, "num_tokens": 655738823.0, "step": 17187 }, { "epoch": 2.1864902684136878, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.570570945739746, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8644136190414429, "num_tokens": 655781457.0, "step": 17188 }, { "epoch": 2.1866174786922783, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.561548233032227, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8688879013061523, "num_tokens": 655824704.0, "step": 17189 }, { "epoch": 2.186744688970869, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.569672584533691, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8830711245536804, "num_tokens": 655860013.0, "step": 17190 }, { "epoch": 2.1868718992494594, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.60712718963623, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.871936559677124, "num_tokens": 655897146.0, "step": 17191 }, { "epoch": 2.18699910952805, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.548978805541992, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8794022798538208, "num_tokens": 655934299.0, "step": 17192 }, { "epoch": 2.1871263198066404, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.569687843322754, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8774460554122925, "num_tokens": 655976554.0, "step": 17193 }, { "epoch": 2.187253530085231, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.581897735595703, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8852950930595398, "num_tokens": 656011756.0, "step": 17194 }, { "epoch": 2.1873807403638215, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.609752655029297, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8710360527038574, "num_tokens": 656054405.0, "step": 17195 }, { "epoch": 2.187507950642412, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.600067138671875, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8779512643814087, "num_tokens": 656094668.0, "step": 17196 }, { "epoch": 2.1876351609210025, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.551356315612793, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8736059069633484, "num_tokens": 656137664.0, "step": 17197 }, { "epoch": 2.187762371199593, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.61121940612793, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8665093183517456, "num_tokens": 656180611.0, "step": 17198 }, { "epoch": 2.1878895814781836, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.552019119262695, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8655300140380859, "num_tokens": 656217789.0, "step": 17199 }, { "epoch": 2.188016791756774, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.578730583190918, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8712361454963684, "num_tokens": 656261940.0, "step": 17200 }, { "epoch": 2.1881440020353646, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.56941032409668, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8835689425468445, "num_tokens": 656296000.0, "step": 17201 }, { "epoch": 2.188271212313955, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.613240242004395, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8840702772140503, "num_tokens": 656340862.0, "step": 17202 }, { "epoch": 2.1883984225925452, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.547449111938477, "learning_rate": 1e-06, "loss": 0.3273, "mean_token_accuracy": 0.8968296051025391, "num_tokens": 656376923.0, "step": 17203 }, { "epoch": 2.188525632871136, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.591060638427734, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8698947429656982, "num_tokens": 656414058.0, "step": 17204 }, { "epoch": 2.1886528431497263, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.567361831665039, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.865504264831543, "num_tokens": 656450077.0, "step": 17205 }, { "epoch": 2.188780053428317, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.611030578613281, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8652773499488831, "num_tokens": 656487258.0, "step": 17206 }, { "epoch": 2.1889072637069074, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.540112495422363, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8881797790527344, "num_tokens": 656525141.0, "step": 17207 }, { "epoch": 2.189034473985498, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.602171897888184, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8766627311706543, "num_tokens": 656561517.0, "step": 17208 }, { "epoch": 2.1891616842640884, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.590437889099121, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8795520067214966, "num_tokens": 656595859.0, "step": 17209 }, { "epoch": 2.189288894542679, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.657670974731445, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8701290488243103, "num_tokens": 656627652.0, "step": 17210 }, { "epoch": 2.1894161048212695, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.592921257019043, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8802623748779297, "num_tokens": 656669951.0, "step": 17211 }, { "epoch": 2.18954331509986, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.530427932739258, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8853644728660583, "num_tokens": 656702216.0, "step": 17212 }, { "epoch": 2.1896705253784505, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.582062721252441, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8638478517532349, "num_tokens": 656744816.0, "step": 17213 }, { "epoch": 2.189797735657041, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.561983108520508, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8701566457748413, "num_tokens": 656785533.0, "step": 17214 }, { "epoch": 2.1899249459356316, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.516924858093262, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8722257018089294, "num_tokens": 656825375.0, "step": 17215 }, { "epoch": 2.190052156214222, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.582935333251953, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8817896842956543, "num_tokens": 656857951.0, "step": 17216 }, { "epoch": 2.1901793664928126, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.557064056396484, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8828185796737671, "num_tokens": 656896911.0, "step": 17217 }, { "epoch": 2.190306576771403, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.566097259521484, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8810059428215027, "num_tokens": 656935723.0, "step": 17218 }, { "epoch": 2.1904337870499937, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.62411117553711, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8854950666427612, "num_tokens": 656970881.0, "step": 17219 }, { "epoch": 2.1905609973285842, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.618795394897461, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8696746230125427, "num_tokens": 657007721.0, "step": 17220 }, { "epoch": 2.1906882076071748, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.60197925567627, "learning_rate": 1e-06, "loss": 0.3477, "mean_token_accuracy": 0.8844772577285767, "num_tokens": 657042359.0, "step": 17221 }, { "epoch": 2.1908154178857653, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.630582809448242, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8782111406326294, "num_tokens": 657084662.0, "step": 17222 }, { "epoch": 2.190942628164356, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.617410659790039, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.875907301902771, "num_tokens": 657122002.0, "step": 17223 }, { "epoch": 2.1910698384429463, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.546341896057129, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8756460547447205, "num_tokens": 657158768.0, "step": 17224 }, { "epoch": 2.191197048721537, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.617423057556152, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8623072504997253, "num_tokens": 657193445.0, "step": 17225 }, { "epoch": 2.1913242590001274, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.556662559509277, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8736752271652222, "num_tokens": 657234922.0, "step": 17226 }, { "epoch": 2.191451469278718, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.590113639831543, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8656294941902161, "num_tokens": 657270774.0, "step": 17227 }, { "epoch": 2.191578679557308, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.558077812194824, "learning_rate": 1e-06, "loss": 0.3458, "mean_token_accuracy": 0.8879560232162476, "num_tokens": 657307046.0, "step": 17228 }, { "epoch": 2.1917058898358985, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.51810073852539, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8642727136611938, "num_tokens": 657352865.0, "step": 17229 }, { "epoch": 2.191833100114489, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.605426788330078, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.854625403881073, "num_tokens": 657393372.0, "step": 17230 }, { "epoch": 2.1919603103930796, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.623735427856445, "learning_rate": 1e-06, "loss": 0.464, "mean_token_accuracy": 0.8541213274002075, "num_tokens": 657432445.0, "step": 17231 }, { "epoch": 2.19208752067167, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.536466598510742, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8726561665534973, "num_tokens": 657469145.0, "step": 17232 }, { "epoch": 2.1922147309502606, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.620136260986328, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8886319398880005, "num_tokens": 657510369.0, "step": 17233 }, { "epoch": 2.192341941228851, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.494843482971191, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8801796436309814, "num_tokens": 657550257.0, "step": 17234 }, { "epoch": 2.1924691515074417, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.52120304107666, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.876082181930542, "num_tokens": 657587095.0, "step": 17235 }, { "epoch": 2.1925963617860322, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.521088600158691, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.878046989440918, "num_tokens": 657619710.0, "step": 17236 }, { "epoch": 2.1927235720646228, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.549386978149414, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8700896501541138, "num_tokens": 657653316.0, "step": 17237 }, { "epoch": 2.1928507823432133, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.589826583862305, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.870349645614624, "num_tokens": 657693570.0, "step": 17238 }, { "epoch": 2.192977992621804, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.52685260772705, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8751494884490967, "num_tokens": 657736781.0, "step": 17239 }, { "epoch": 2.1931052029003943, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.538344383239746, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8660518527030945, "num_tokens": 657775603.0, "step": 17240 }, { "epoch": 2.193232413178985, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.576079368591309, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8716082572937012, "num_tokens": 657809065.0, "step": 17241 }, { "epoch": 2.1933596234575754, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.60999870300293, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8721926212310791, "num_tokens": 657851369.0, "step": 17242 }, { "epoch": 2.193486833736166, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.588829040527344, "learning_rate": 1e-06, "loss": 0.3378, "mean_token_accuracy": 0.8949995040893555, "num_tokens": 657889008.0, "step": 17243 }, { "epoch": 2.1936140440147565, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.575345993041992, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8727648258209229, "num_tokens": 657925867.0, "step": 17244 }, { "epoch": 2.193741254293347, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.610400199890137, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8647664189338684, "num_tokens": 657959502.0, "step": 17245 }, { "epoch": 2.1938684645719375, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.532394409179688, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8772412538528442, "num_tokens": 657999063.0, "step": 17246 }, { "epoch": 2.193995674850528, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.622243881225586, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8642624616622925, "num_tokens": 658033016.0, "step": 17247 }, { "epoch": 2.1941228851291186, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.598026275634766, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8802284002304077, "num_tokens": 658071259.0, "step": 17248 }, { "epoch": 2.194250095407709, "ewc_loss": 0.035400390625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 13.570234298706055, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8759837746620178, "num_tokens": 658103190.0, "step": 17249 }, { "epoch": 2.1943773056862996, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.654326438903809, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8662867546081543, "num_tokens": 658143346.0, "step": 17250 }, { "epoch": 2.19450451596489, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.536810874938965, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8829503059387207, "num_tokens": 658181747.0, "step": 17251 }, { "epoch": 2.1946317262434807, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.594964027404785, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8750149011611938, "num_tokens": 658221770.0, "step": 17252 }, { "epoch": 2.1947589365220708, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.56180191040039, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8781826496124268, "num_tokens": 658260157.0, "step": 17253 }, { "epoch": 2.1948861468006613, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.597793579101562, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8637070059776306, "num_tokens": 658301815.0, "step": 17254 }, { "epoch": 2.195013357079252, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.619776725769043, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8759433031082153, "num_tokens": 658340732.0, "step": 17255 }, { "epoch": 2.1951405673578424, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.573187828063965, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8689267635345459, "num_tokens": 658380569.0, "step": 17256 }, { "epoch": 2.195267777636433, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.658952713012695, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8857352137565613, "num_tokens": 658415842.0, "step": 17257 }, { "epoch": 2.1953949879150234, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.586746215820312, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8785425424575806, "num_tokens": 658452861.0, "step": 17258 }, { "epoch": 2.195522198193614, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.65380859375, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8774333000183105, "num_tokens": 658487019.0, "step": 17259 }, { "epoch": 2.1956494084722045, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.607647895812988, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8723632097244263, "num_tokens": 658529290.0, "step": 17260 }, { "epoch": 2.195776618750795, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.603434562683105, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8690622448921204, "num_tokens": 658572118.0, "step": 17261 }, { "epoch": 2.1959038290293855, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.614797592163086, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8707839250564575, "num_tokens": 658615972.0, "step": 17262 }, { "epoch": 2.196031039307976, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.62888240814209, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.881348729133606, "num_tokens": 658653848.0, "step": 17263 }, { "epoch": 2.1961582495865666, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.650303840637207, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8737637996673584, "num_tokens": 658686247.0, "step": 17264 }, { "epoch": 2.196285459865157, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.649513244628906, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8716262578964233, "num_tokens": 658725787.0, "step": 17265 }, { "epoch": 2.1964126701437476, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.604562759399414, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8705883026123047, "num_tokens": 658765957.0, "step": 17266 }, { "epoch": 2.196539880422338, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.585322380065918, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.883335292339325, "num_tokens": 658800996.0, "step": 17267 }, { "epoch": 2.1966670907009287, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.567985534667969, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8716894388198853, "num_tokens": 658833779.0, "step": 17268 }, { "epoch": 2.196794300979519, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.54403305053711, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8775082230567932, "num_tokens": 658874425.0, "step": 17269 }, { "epoch": 2.1969215112581097, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.61961555480957, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8825997710227966, "num_tokens": 658912007.0, "step": 17270 }, { "epoch": 2.1970487215367003, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.561488151550293, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8651571273803711, "num_tokens": 658956775.0, "step": 17271 }, { "epoch": 2.197175931815291, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.681509017944336, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8866227269172668, "num_tokens": 658997132.0, "step": 17272 }, { "epoch": 2.1973031420938813, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.575807571411133, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.880959153175354, "num_tokens": 659040437.0, "step": 17273 }, { "epoch": 2.197430352372472, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.631796836853027, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.889204740524292, "num_tokens": 659074461.0, "step": 17274 }, { "epoch": 2.1975575626510624, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.609341621398926, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8712713718414307, "num_tokens": 659113818.0, "step": 17275 }, { "epoch": 2.197684772929653, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.61697769165039, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8618336319923401, "num_tokens": 659149297.0, "step": 17276 }, { "epoch": 2.1978119832082434, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.604020118713379, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8770350813865662, "num_tokens": 659187500.0, "step": 17277 }, { "epoch": 2.1979391934868335, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.565367698669434, "learning_rate": 1e-06, "loss": 0.4822, "mean_token_accuracy": 0.848383903503418, "num_tokens": 659228624.0, "step": 17278 }, { "epoch": 2.198066403765424, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.602745056152344, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8811386227607727, "num_tokens": 659270317.0, "step": 17279 }, { "epoch": 2.1981936140440146, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.573718070983887, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8735986948013306, "num_tokens": 659302133.0, "step": 17280 }, { "epoch": 2.198320824322605, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.597530364990234, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8796991109848022, "num_tokens": 659342469.0, "step": 17281 }, { "epoch": 2.1984480346011956, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.61038875579834, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8851738572120667, "num_tokens": 659373786.0, "step": 17282 }, { "epoch": 2.198575244879786, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.584568977355957, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8678290843963623, "num_tokens": 659412498.0, "step": 17283 }, { "epoch": 2.1987024551583767, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.619729042053223, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8591513633728027, "num_tokens": 659446415.0, "step": 17284 }, { "epoch": 2.1988296654369672, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.591520309448242, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8661697506904602, "num_tokens": 659488824.0, "step": 17285 }, { "epoch": 2.1989568757155578, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.61397933959961, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8781912326812744, "num_tokens": 659523036.0, "step": 17286 }, { "epoch": 2.1990840859941483, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.62977409362793, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8680978417396545, "num_tokens": 659562780.0, "step": 17287 }, { "epoch": 2.199211296272739, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.637079238891602, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8809447288513184, "num_tokens": 659599364.0, "step": 17288 }, { "epoch": 2.1993385065513293, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.583291053771973, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8673803806304932, "num_tokens": 659636275.0, "step": 17289 }, { "epoch": 2.19946571682992, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.607596397399902, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8740025162696838, "num_tokens": 659670315.0, "step": 17290 }, { "epoch": 2.1995929271085104, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.60054874420166, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8701215982437134, "num_tokens": 659710968.0, "step": 17291 }, { "epoch": 2.199720137387101, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.64157772064209, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.882548987865448, "num_tokens": 659751334.0, "step": 17292 }, { "epoch": 2.1998473476656915, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.643625259399414, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.875097930431366, "num_tokens": 659793867.0, "step": 17293 }, { "epoch": 2.199974557944282, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.553693771362305, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8746188879013062, "num_tokens": 659838724.0, "step": 17294 }, { "epoch": 2.2001017682228725, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.653076171875, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8726130127906799, "num_tokens": 659876589.0, "step": 17295 }, { "epoch": 2.200228978501463, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.67220401763916, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8651673793792725, "num_tokens": 659914864.0, "step": 17296 }, { "epoch": 2.2003561887800536, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.604811668395996, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8806955814361572, "num_tokens": 659952482.0, "step": 17297 }, { "epoch": 2.200483399058644, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.632406234741211, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8746322393417358, "num_tokens": 659985192.0, "step": 17298 }, { "epoch": 2.2006106093372346, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.612570762634277, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8814568519592285, "num_tokens": 660019505.0, "step": 17299 }, { "epoch": 2.200737819615825, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.608050346374512, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.867688000202179, "num_tokens": 660054942.0, "step": 17300 }, { "epoch": 2.2008650298944152, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.60374641418457, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8739954829216003, "num_tokens": 660092836.0, "step": 17301 }, { "epoch": 2.200992240173006, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.584298133850098, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8685157895088196, "num_tokens": 660129419.0, "step": 17302 }, { "epoch": 2.2011194504515963, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.594557762145996, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8772033452987671, "num_tokens": 660168116.0, "step": 17303 }, { "epoch": 2.201246660730187, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.547492027282715, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8687896728515625, "num_tokens": 660208006.0, "step": 17304 }, { "epoch": 2.2013738710087773, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.566490173339844, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8733553886413574, "num_tokens": 660254046.0, "step": 17305 }, { "epoch": 2.201501081287368, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.619552612304688, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.872706949710846, "num_tokens": 660289321.0, "step": 17306 }, { "epoch": 2.2016282915659584, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.603117942810059, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8714457154273987, "num_tokens": 660325052.0, "step": 17307 }, { "epoch": 2.201755501844549, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.618877410888672, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8653702735900879, "num_tokens": 660357942.0, "step": 17308 }, { "epoch": 2.2018827121231395, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.610579490661621, "learning_rate": 1e-06, "loss": 0.4782, "mean_token_accuracy": 0.8482866883277893, "num_tokens": 660390654.0, "step": 17309 }, { "epoch": 2.20200992240173, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.581629753112793, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8749716281890869, "num_tokens": 660432020.0, "step": 17310 }, { "epoch": 2.2021371326803205, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.602227210998535, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8740720152854919, "num_tokens": 660465051.0, "step": 17311 }, { "epoch": 2.202264342958911, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.604677200317383, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8703676462173462, "num_tokens": 660504967.0, "step": 17312 }, { "epoch": 2.2023915532375016, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.574455261230469, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8636569976806641, "num_tokens": 660543826.0, "step": 17313 }, { "epoch": 2.202518763516092, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.629251480102539, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8700258731842041, "num_tokens": 660583074.0, "step": 17314 }, { "epoch": 2.2026459737946826, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.589736938476562, "learning_rate": 1e-06, "loss": 0.344, "mean_token_accuracy": 0.8895959854125977, "num_tokens": 660619974.0, "step": 17315 }, { "epoch": 2.202773184073273, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.581271171569824, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8797354698181152, "num_tokens": 660654005.0, "step": 17316 }, { "epoch": 2.2029003943518637, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.615221977233887, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8628829717636108, "num_tokens": 660689179.0, "step": 17317 }, { "epoch": 2.203027604630454, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.618876457214355, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8622077703475952, "num_tokens": 660724923.0, "step": 17318 }, { "epoch": 2.2031548149090447, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.580605506896973, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8773953914642334, "num_tokens": 660760763.0, "step": 17319 }, { "epoch": 2.2032820251876353, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.630363464355469, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8738218545913696, "num_tokens": 660795520.0, "step": 17320 }, { "epoch": 2.203409235466226, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.522677421569824, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8839662075042725, "num_tokens": 660834858.0, "step": 17321 }, { "epoch": 2.2035364457448163, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.545223236083984, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8824237585067749, "num_tokens": 660870319.0, "step": 17322 }, { "epoch": 2.203663656023407, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.64087963104248, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8608928918838501, "num_tokens": 660907313.0, "step": 17323 }, { "epoch": 2.2037908663019974, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.521649360656738, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.8886198997497559, "num_tokens": 660948087.0, "step": 17324 }, { "epoch": 2.203918076580588, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.643373489379883, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8794819116592407, "num_tokens": 660984984.0, "step": 17325 }, { "epoch": 2.204045286859178, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.640812873840332, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8705785274505615, "num_tokens": 661022898.0, "step": 17326 }, { "epoch": 2.2041724971377685, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.575051307678223, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8833746910095215, "num_tokens": 661058645.0, "step": 17327 }, { "epoch": 2.204299707416359, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.556297302246094, "learning_rate": 1e-06, "loss": 0.3395, "mean_token_accuracy": 0.8958856463432312, "num_tokens": 661096954.0, "step": 17328 }, { "epoch": 2.2044269176949496, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.564346313476562, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8631223440170288, "num_tokens": 661136947.0, "step": 17329 }, { "epoch": 2.20455412797354, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.62801742553711, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8755272626876831, "num_tokens": 661175376.0, "step": 17330 }, { "epoch": 2.2046813382521306, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.52863597869873, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8663777112960815, "num_tokens": 661213238.0, "step": 17331 }, { "epoch": 2.204808548530721, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.562944412231445, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8890166282653809, "num_tokens": 661246512.0, "step": 17332 }, { "epoch": 2.2049357588093117, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.580482482910156, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8844296932220459, "num_tokens": 661288272.0, "step": 17333 }, { "epoch": 2.2050629690879022, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.54783821105957, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.879772424697876, "num_tokens": 661328705.0, "step": 17334 }, { "epoch": 2.2051901793664928, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.586088180541992, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8783059120178223, "num_tokens": 661369877.0, "step": 17335 }, { "epoch": 2.2053173896450833, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.58852481842041, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.877011775970459, "num_tokens": 661405209.0, "step": 17336 }, { "epoch": 2.205444599923674, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.643169403076172, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8783656358718872, "num_tokens": 661441083.0, "step": 17337 }, { "epoch": 2.2055718102022643, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.769755363464355, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8762226104736328, "num_tokens": 661476938.0, "step": 17338 }, { "epoch": 2.205699020480855, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.548173904418945, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8665715456008911, "num_tokens": 661519281.0, "step": 17339 }, { "epoch": 2.2058262307594454, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.539029121398926, "learning_rate": 1e-06, "loss": 0.3455, "mean_token_accuracy": 0.8894981145858765, "num_tokens": 661555880.0, "step": 17340 }, { "epoch": 2.205953441038036, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.64769172668457, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8659995198249817, "num_tokens": 661595937.0, "step": 17341 }, { "epoch": 2.2060806513166265, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.594245910644531, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8608906269073486, "num_tokens": 661636868.0, "step": 17342 }, { "epoch": 2.206207861595217, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.570359230041504, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8678261041641235, "num_tokens": 661681503.0, "step": 17343 }, { "epoch": 2.2063350718738075, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.633259773254395, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8780415058135986, "num_tokens": 661718438.0, "step": 17344 }, { "epoch": 2.206462282152398, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.605424880981445, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8543134331703186, "num_tokens": 661759048.0, "step": 17345 }, { "epoch": 2.2065894924309886, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.565351486206055, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8846834301948547, "num_tokens": 661803131.0, "step": 17346 }, { "epoch": 2.206716702709579, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.636598587036133, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8741441369056702, "num_tokens": 661838139.0, "step": 17347 }, { "epoch": 2.2068439129881696, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.569580078125, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8680026531219482, "num_tokens": 661878568.0, "step": 17348 }, { "epoch": 2.20697112326676, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.640646934509277, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8688563704490662, "num_tokens": 661909110.0, "step": 17349 }, { "epoch": 2.2070983335453507, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.605944633483887, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8871132731437683, "num_tokens": 661944472.0, "step": 17350 }, { "epoch": 2.2072255438239408, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.642396926879883, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8860854506492615, "num_tokens": 661985865.0, "step": 17351 }, { "epoch": 2.2073527541025313, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.595970153808594, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8689501285552979, "num_tokens": 662019460.0, "step": 17352 }, { "epoch": 2.207479964381122, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.56615161895752, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8807865977287292, "num_tokens": 662059109.0, "step": 17353 }, { "epoch": 2.2076071746597123, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.700124740600586, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8675428628921509, "num_tokens": 662096040.0, "step": 17354 }, { "epoch": 2.207734384938303, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.573554039001465, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8791515231132507, "num_tokens": 662139351.0, "step": 17355 }, { "epoch": 2.2078615952168934, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.606914520263672, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8751147389411926, "num_tokens": 662177353.0, "step": 17356 }, { "epoch": 2.207988805495484, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.587176322937012, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8641239404678345, "num_tokens": 662209519.0, "step": 17357 }, { "epoch": 2.2081160157740745, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.625732421875, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8555952906608582, "num_tokens": 662251311.0, "step": 17358 }, { "epoch": 2.208243226052665, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.674628257751465, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8813173770904541, "num_tokens": 662282770.0, "step": 17359 }, { "epoch": 2.2083704363312555, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.585556030273438, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8705275058746338, "num_tokens": 662326770.0, "step": 17360 }, { "epoch": 2.208497646609846, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.571028709411621, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8736835718154907, "num_tokens": 662366811.0, "step": 17361 }, { "epoch": 2.2086248568884366, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.603610038757324, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8700017929077148, "num_tokens": 662404517.0, "step": 17362 }, { "epoch": 2.208752067167027, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.643771171569824, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8800191283226013, "num_tokens": 662442091.0, "step": 17363 }, { "epoch": 2.2088792774456176, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.576781272888184, "learning_rate": 1e-06, "loss": 0.4719, "mean_token_accuracy": 0.8492192029953003, "num_tokens": 662483343.0, "step": 17364 }, { "epoch": 2.209006487724208, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.612106323242188, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.868106484413147, "num_tokens": 662521994.0, "step": 17365 }, { "epoch": 2.2091336980027987, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.581828117370605, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8702956438064575, "num_tokens": 662562737.0, "step": 17366 }, { "epoch": 2.209260908281389, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.616628646850586, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8708113431930542, "num_tokens": 662596927.0, "step": 17367 }, { "epoch": 2.2093881185599797, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.631978988647461, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8816201686859131, "num_tokens": 662632831.0, "step": 17368 }, { "epoch": 2.2095153288385703, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.596394538879395, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8707951903343201, "num_tokens": 662675554.0, "step": 17369 }, { "epoch": 2.209642539117161, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.608563423156738, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8852474689483643, "num_tokens": 662720248.0, "step": 17370 }, { "epoch": 2.2097697493957513, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.552229881286621, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8735685348510742, "num_tokens": 662761388.0, "step": 17371 }, { "epoch": 2.209896959674342, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.632059097290039, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8727474808692932, "num_tokens": 662798007.0, "step": 17372 }, { "epoch": 2.2100241699529324, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.615274429321289, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8718124032020569, "num_tokens": 662836042.0, "step": 17373 }, { "epoch": 2.210151380231523, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.634658813476562, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8651163578033447, "num_tokens": 662869465.0, "step": 17374 }, { "epoch": 2.2102785905101134, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.580355644226074, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8655561208724976, "num_tokens": 662909440.0, "step": 17375 }, { "epoch": 2.2104058007887035, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.656135559082031, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8703340291976929, "num_tokens": 662949637.0, "step": 17376 }, { "epoch": 2.210533011067294, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.668237686157227, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.871820330619812, "num_tokens": 662987679.0, "step": 17377 }, { "epoch": 2.2106602213458846, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.618796348571777, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8547731637954712, "num_tokens": 663027651.0, "step": 17378 }, { "epoch": 2.210787431624475, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.684656143188477, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8863579630851746, "num_tokens": 663063708.0, "step": 17379 }, { "epoch": 2.2109146419030656, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.65247917175293, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8662709593772888, "num_tokens": 663097255.0, "step": 17380 }, { "epoch": 2.211041852181656, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.68376350402832, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8703043460845947, "num_tokens": 663134496.0, "step": 17381 }, { "epoch": 2.2111690624602467, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.59781265258789, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8820167183876038, "num_tokens": 663170278.0, "step": 17382 }, { "epoch": 2.211296272738837, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.71495246887207, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8710007667541504, "num_tokens": 663205407.0, "step": 17383 }, { "epoch": 2.2114234830174277, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.603965759277344, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.878057599067688, "num_tokens": 663243166.0, "step": 17384 }, { "epoch": 2.2115506932960183, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.63314437866211, "learning_rate": 1e-06, "loss": 0.3336, "mean_token_accuracy": 0.8950239419937134, "num_tokens": 663282267.0, "step": 17385 }, { "epoch": 2.211677903574609, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.655309677124023, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8657311797142029, "num_tokens": 663318620.0, "step": 17386 }, { "epoch": 2.2118051138531993, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.607327461242676, "learning_rate": 1e-06, "loss": 0.3413, "mean_token_accuracy": 0.891382098197937, "num_tokens": 663353256.0, "step": 17387 }, { "epoch": 2.21193232413179, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.640317916870117, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.877036988735199, "num_tokens": 663386691.0, "step": 17388 }, { "epoch": 2.2120595344103804, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.649874687194824, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8703004121780396, "num_tokens": 663416550.0, "step": 17389 }, { "epoch": 2.212186744688971, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.67695426940918, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8753334879875183, "num_tokens": 663464912.0, "step": 17390 }, { "epoch": 2.2123139549675614, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.618086814880371, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.864936888217926, "num_tokens": 663503786.0, "step": 17391 }, { "epoch": 2.212441165246152, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.648970603942871, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8815853595733643, "num_tokens": 663540579.0, "step": 17392 }, { "epoch": 2.2125683755247425, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.618661880493164, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8757643699645996, "num_tokens": 663577284.0, "step": 17393 }, { "epoch": 2.212695585803333, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.625977516174316, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8696218132972717, "num_tokens": 663613118.0, "step": 17394 }, { "epoch": 2.2128227960819236, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.65177059173584, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.87729811668396, "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": 13.609747886657715, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8565446734428406, "num_tokens": 663694381.0, "step": 17396 }, { "epoch": 2.2130772166391046, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.600911140441895, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8744106292724609, "num_tokens": 663734686.0, "step": 17397 }, { "epoch": 2.213204426917695, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.681082725524902, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8657446503639221, "num_tokens": 663768764.0, "step": 17398 }, { "epoch": 2.2133316371962852, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.576567649841309, "learning_rate": 1e-06, "loss": 0.3388, "mean_token_accuracy": 0.8936249613761902, "num_tokens": 663807650.0, "step": 17399 }, { "epoch": 2.213458847474876, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.706718444824219, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.871873140335083, "num_tokens": 663849308.0, "step": 17400 }, { "epoch": 2.2135860577534663, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.621798515319824, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8602344989776611, "num_tokens": 663888917.0, "step": 17401 }, { "epoch": 2.213713268032057, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.694836616516113, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8744356632232666, "num_tokens": 663921048.0, "step": 17402 }, { "epoch": 2.2138404783106473, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.589499473571777, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8786731362342834, "num_tokens": 663963433.0, "step": 17403 }, { "epoch": 2.213967688589238, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.682634353637695, "learning_rate": 1e-06, "loss": 0.4566, "mean_token_accuracy": 0.8569483757019043, "num_tokens": 664003294.0, "step": 17404 }, { "epoch": 2.2140948988678284, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.628412246704102, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.861698567867279, "num_tokens": 664043161.0, "step": 17405 }, { "epoch": 2.214222109146419, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.630452156066895, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8850334882736206, "num_tokens": 664082457.0, "step": 17406 }, { "epoch": 2.2143493194250095, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.664456367492676, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8831943869590759, "num_tokens": 664125999.0, "step": 17407 }, { "epoch": 2.2144765297036, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.608163833618164, "learning_rate": 1e-06, "loss": 0.3789, "mean_token_accuracy": 0.8790940046310425, "num_tokens": 664167256.0, "step": 17408 }, { "epoch": 2.2146037399821905, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.695030212402344, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8698012828826904, "num_tokens": 664201887.0, "step": 17409 }, { "epoch": 2.214730950260781, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.663651466369629, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8786594867706299, "num_tokens": 664238986.0, "step": 17410 }, { "epoch": 2.2148581605393716, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.64894962310791, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8699559569358826, "num_tokens": 664283436.0, "step": 17411 }, { "epoch": 2.214985370817962, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.675824165344238, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8672882914543152, "num_tokens": 664325196.0, "step": 17412 }, { "epoch": 2.2151125810965526, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.600271224975586, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8647798895835876, "num_tokens": 664365722.0, "step": 17413 }, { "epoch": 2.215239791375143, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.630672454833984, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8644024133682251, "num_tokens": 664411570.0, "step": 17414 }, { "epoch": 2.2153670016537337, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.608421325683594, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8764897584915161, "num_tokens": 664453545.0, "step": 17415 }, { "epoch": 2.215494211932324, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.678285598754883, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8744738101959229, "num_tokens": 664489882.0, "step": 17416 }, { "epoch": 2.2156214222109147, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.676992416381836, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8708082437515259, "num_tokens": 664528720.0, "step": 17417 }, { "epoch": 2.2157486324895053, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.631053924560547, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8641950488090515, "num_tokens": 664569645.0, "step": 17418 }, { "epoch": 2.215875842768096, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.677520751953125, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8680970668792725, "num_tokens": 664612033.0, "step": 17419 }, { "epoch": 2.2160030530466863, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.632694244384766, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8813446760177612, "num_tokens": 664651408.0, "step": 17420 }, { "epoch": 2.216130263325277, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.643023490905762, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8583593964576721, "num_tokens": 664695764.0, "step": 17421 }, { "epoch": 2.2162574736038674, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.625799179077148, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8765227794647217, "num_tokens": 664735528.0, "step": 17422 }, { "epoch": 2.216384683882458, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.609004020690918, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8655937910079956, "num_tokens": 664775612.0, "step": 17423 }, { "epoch": 2.216511894161048, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.62752914428711, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.869101881980896, "num_tokens": 664812485.0, "step": 17424 }, { "epoch": 2.2166391044396385, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.64195728302002, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8585836291313171, "num_tokens": 664853166.0, "step": 17425 }, { "epoch": 2.216766314718229, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.640768051147461, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8631453514099121, "num_tokens": 664888778.0, "step": 17426 }, { "epoch": 2.2168935249968196, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.614645004272461, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8714981079101562, "num_tokens": 664933783.0, "step": 17427 }, { "epoch": 2.21702073527541, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.646280288696289, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8566306829452515, "num_tokens": 664965228.0, "step": 17428 }, { "epoch": 2.2171479455540006, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.686758041381836, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8691888451576233, "num_tokens": 664996691.0, "step": 17429 }, { "epoch": 2.217275155832591, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.601449966430664, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8836885094642639, "num_tokens": 665032006.0, "step": 17430 }, { "epoch": 2.2174023661111817, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.607207298278809, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8628982305526733, "num_tokens": 665071819.0, "step": 17431 }, { "epoch": 2.217529576389772, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.600617408752441, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8615228533744812, "num_tokens": 665118883.0, "step": 17432 }, { "epoch": 2.2176567866683627, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.650487899780273, "learning_rate": 1e-06, "loss": 0.4663, "mean_token_accuracy": 0.8546355366706848, "num_tokens": 665154926.0, "step": 17433 }, { "epoch": 2.2177839969469533, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.655692100524902, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8809673190116882, "num_tokens": 665189910.0, "step": 17434 }, { "epoch": 2.217911207225544, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.623311996459961, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8797029256820679, "num_tokens": 665235849.0, "step": 17435 }, { "epoch": 2.2180384175041343, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.64831256866455, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8636678457260132, "num_tokens": 665274231.0, "step": 17436 }, { "epoch": 2.218165627782725, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.606999397277832, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8910778760910034, "num_tokens": 665309057.0, "step": 17437 }, { "epoch": 2.2182928380613154, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.711225509643555, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8711153864860535, "num_tokens": 665337031.0, "step": 17438 }, { "epoch": 2.218420048339906, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.610015869140625, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8725374937057495, "num_tokens": 665376557.0, "step": 17439 }, { "epoch": 2.2185472586184964, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.676630973815918, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8701421022415161, "num_tokens": 665419928.0, "step": 17440 }, { "epoch": 2.218674468897087, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.596396446228027, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8789592385292053, "num_tokens": 665458827.0, "step": 17441 }, { "epoch": 2.2188016791756775, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.640424728393555, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8745195269584656, "num_tokens": 665498623.0, "step": 17442 }, { "epoch": 2.218928889454268, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.669336318969727, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8680248260498047, "num_tokens": 665536103.0, "step": 17443 }, { "epoch": 2.2190560997328586, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.681442260742188, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.87042635679245, "num_tokens": 665572089.0, "step": 17444 }, { "epoch": 2.219183310011449, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.63449764251709, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8649926781654358, "num_tokens": 665608088.0, "step": 17445 }, { "epoch": 2.2193105202900396, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.583555221557617, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8818209767341614, "num_tokens": 665641883.0, "step": 17446 }, { "epoch": 2.21943773056863, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.666825294494629, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8771848678588867, "num_tokens": 665682211.0, "step": 17447 }, { "epoch": 2.2195649408472207, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.65377140045166, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8774024248123169, "num_tokens": 665726182.0, "step": 17448 }, { "epoch": 2.2196921511258108, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.633259773254395, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8762013912200928, "num_tokens": 665760404.0, "step": 17449 }, { "epoch": 2.2198193614044013, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.63074779510498, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8726539611816406, "num_tokens": 665798235.0, "step": 17450 }, { "epoch": 2.219946571682992, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.656525611877441, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8730928301811218, "num_tokens": 665840118.0, "step": 17451 }, { "epoch": 2.2200737819615823, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.6299467086792, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8721368312835693, "num_tokens": 665878498.0, "step": 17452 }, { "epoch": 2.220200992240173, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.613149642944336, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8695428371429443, "num_tokens": 665924179.0, "step": 17453 }, { "epoch": 2.2203282025187634, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.639687538146973, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8859797716140747, "num_tokens": 665963160.0, "step": 17454 }, { "epoch": 2.220455412797354, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.743827819824219, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8818930983543396, "num_tokens": 665998854.0, "step": 17455 }, { "epoch": 2.2205826230759445, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.584705352783203, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8808761239051819, "num_tokens": 666036420.0, "step": 17456 }, { "epoch": 2.220709833354535, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.645819664001465, "learning_rate": 1e-06, "loss": 0.4677, "mean_token_accuracy": 0.8544743061065674, "num_tokens": 666076692.0, "step": 17457 }, { "epoch": 2.2208370436331255, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.649697303771973, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8597103357315063, "num_tokens": 666118009.0, "step": 17458 }, { "epoch": 2.220964253911716, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.669901847839355, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.863482654094696, "num_tokens": 666158132.0, "step": 17459 }, { "epoch": 2.2210914641903066, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.679773330688477, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8669864535331726, "num_tokens": 666187195.0, "step": 17460 }, { "epoch": 2.221218674468897, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.636075019836426, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8779193162918091, "num_tokens": 666227790.0, "step": 17461 }, { "epoch": 2.2213458847474876, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.608996391296387, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8665339946746826, "num_tokens": 666268517.0, "step": 17462 }, { "epoch": 2.221473095026078, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.6444091796875, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8893970251083374, "num_tokens": 666313727.0, "step": 17463 }, { "epoch": 2.2216003053046687, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.62962532043457, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8758609294891357, "num_tokens": 666352713.0, "step": 17464 }, { "epoch": 2.221727515583259, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.625958442687988, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8800936341285706, "num_tokens": 666395065.0, "step": 17465 }, { "epoch": 2.2218547258618497, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.59615707397461, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8749863505363464, "num_tokens": 666428019.0, "step": 17466 }, { "epoch": 2.2219819361404403, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.690378189086914, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.871576189994812, "num_tokens": 666466830.0, "step": 17467 }, { "epoch": 2.222109146419031, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.575037956237793, "learning_rate": 1e-06, "loss": 0.3447, "mean_token_accuracy": 0.8894234895706177, "num_tokens": 666501028.0, "step": 17468 }, { "epoch": 2.2222363566976213, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.685657501220703, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8856058120727539, "num_tokens": 666538554.0, "step": 17469 }, { "epoch": 2.222363566976212, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.602890968322754, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8720412254333496, "num_tokens": 666577019.0, "step": 17470 }, { "epoch": 2.2224907772548024, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.664033889770508, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8869677782058716, "num_tokens": 666616597.0, "step": 17471 }, { "epoch": 2.222617987533393, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.632638931274414, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8767764568328857, "num_tokens": 666651523.0, "step": 17472 }, { "epoch": 2.2227451978119834, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.613381385803223, "learning_rate": 1e-06, "loss": 0.4441, "mean_token_accuracy": 0.863821268081665, "num_tokens": 666689270.0, "step": 17473 }, { "epoch": 2.2228724080905735, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.71588134765625, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8698478937149048, "num_tokens": 666726836.0, "step": 17474 }, { "epoch": 2.222999618369164, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.60233211517334, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8702676296234131, "num_tokens": 666761207.0, "step": 17475 }, { "epoch": 2.2231268286477546, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.717474937438965, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8810730576515198, "num_tokens": 666800293.0, "step": 17476 }, { "epoch": 2.223254038926345, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.622599601745605, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8815757036209106, "num_tokens": 666843046.0, "step": 17477 }, { "epoch": 2.2233812492049356, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.638338088989258, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8868211507797241, "num_tokens": 666885840.0, "step": 17478 }, { "epoch": 2.223508459483526, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.645085334777832, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8738698959350586, "num_tokens": 666925654.0, "step": 17479 }, { "epoch": 2.2236356697621167, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.631340026855469, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8746947050094604, "num_tokens": 666968098.0, "step": 17480 }, { "epoch": 2.223762880040707, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.648893356323242, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8837242126464844, "num_tokens": 667010349.0, "step": 17481 }, { "epoch": 2.2238900903192977, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.617115020751953, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8644874095916748, "num_tokens": 667047086.0, "step": 17482 }, { "epoch": 2.2240173005978883, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.6915864944458, "learning_rate": 1e-06, "loss": 0.4538, "mean_token_accuracy": 0.8566957712173462, "num_tokens": 667083955.0, "step": 17483 }, { "epoch": 2.224144510876479, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.630904197692871, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8800182342529297, "num_tokens": 667119751.0, "step": 17484 }, { "epoch": 2.2242717211550693, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.749812126159668, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8691980838775635, "num_tokens": 667150980.0, "step": 17485 }, { "epoch": 2.22439893143366, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.627043724060059, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8698906898498535, "num_tokens": 667185198.0, "step": 17486 }, { "epoch": 2.2245261417122504, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.635998725891113, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8762797117233276, "num_tokens": 667225286.0, "step": 17487 }, { "epoch": 2.224653351990841, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.688430786132812, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8671643733978271, "num_tokens": 667266294.0, "step": 17488 }, { "epoch": 2.2247805622694314, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.633356094360352, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8706183433532715, "num_tokens": 667302430.0, "step": 17489 }, { "epoch": 2.224907772548022, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.628547668457031, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8674970865249634, "num_tokens": 667342881.0, "step": 17490 }, { "epoch": 2.2250349828266125, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.686219215393066, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8751668334007263, "num_tokens": 667376122.0, "step": 17491 }, { "epoch": 2.225162193105203, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.634687423706055, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8660327196121216, "num_tokens": 667415120.0, "step": 17492 }, { "epoch": 2.2252894033837936, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.637799263000488, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.872943639755249, "num_tokens": 667451014.0, "step": 17493 }, { "epoch": 2.225416613662384, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.672082901000977, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8640279769897461, "num_tokens": 667485600.0, "step": 17494 }, { "epoch": 2.2255438239409746, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.666739463806152, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.868881106376648, "num_tokens": 667527561.0, "step": 17495 }, { "epoch": 2.225671034219565, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.632994651794434, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8809996247291565, "num_tokens": 667569677.0, "step": 17496 }, { "epoch": 2.225798244498155, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.670975685119629, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8729599714279175, "num_tokens": 667605954.0, "step": 17497 }, { "epoch": 2.225925454776746, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.62177848815918, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8762716054916382, "num_tokens": 667645082.0, "step": 17498 }, { "epoch": 2.2260526650553363, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.687694549560547, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8699597120285034, "num_tokens": 667681390.0, "step": 17499 }, { "epoch": 2.226179875333927, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.625487327575684, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8629863262176514, "num_tokens": 667722772.0, "step": 17500 }, { "epoch": 2.2263070856125173, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.684582710266113, "learning_rate": 1e-06, "loss": 0.4758, "mean_token_accuracy": 0.8490922451019287, "num_tokens": 667758647.0, "step": 17501 }, { "epoch": 2.226434295891108, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.643447875976562, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8730829954147339, "num_tokens": 667790625.0, "step": 17502 }, { "epoch": 2.2265615061696984, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.65325927734375, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8626236319541931, "num_tokens": 667823724.0, "step": 17503 }, { "epoch": 2.226688716448289, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.589357376098633, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8862800002098083, "num_tokens": 667862548.0, "step": 17504 }, { "epoch": 2.2268159267268794, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.624383926391602, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8660633563995361, "num_tokens": 667905994.0, "step": 17505 }, { "epoch": 2.22694313700547, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.665637969970703, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8666114807128906, "num_tokens": 667945087.0, "step": 17506 }, { "epoch": 2.2270703472840605, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.694005012512207, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8822104930877686, "num_tokens": 667984736.0, "step": 17507 }, { "epoch": 2.227197557562651, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.666777610778809, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8677756786346436, "num_tokens": 668024020.0, "step": 17508 }, { "epoch": 2.2273247678412416, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.634387016296387, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8687782287597656, "num_tokens": 668059567.0, "step": 17509 }, { "epoch": 2.227451978119832, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.663738250732422, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8792838454246521, "num_tokens": 668099901.0, "step": 17510 }, { "epoch": 2.2275791883984226, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.667173385620117, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8871462941169739, "num_tokens": 668135974.0, "step": 17511 }, { "epoch": 2.227706398677013, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.639313697814941, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8756633996963501, "num_tokens": 668179200.0, "step": 17512 }, { "epoch": 2.2278336089556037, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.633378982543945, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.851630687713623, "num_tokens": 668219683.0, "step": 17513 }, { "epoch": 2.227960819234194, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.672325134277344, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8752305507659912, "num_tokens": 668253899.0, "step": 17514 }, { "epoch": 2.2280880295127847, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.72142219543457, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8676652908325195, "num_tokens": 668294140.0, "step": 17515 }, { "epoch": 2.2282152397913753, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.556286811828613, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8786864280700684, "num_tokens": 668328442.0, "step": 17516 }, { "epoch": 2.228342450069966, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.66763973236084, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.87568199634552, "num_tokens": 668367903.0, "step": 17517 }, { "epoch": 2.2284696603485563, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.654865264892578, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8852746486663818, "num_tokens": 668409889.0, "step": 17518 }, { "epoch": 2.228596870627147, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.668954849243164, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8771213889122009, "num_tokens": 668451061.0, "step": 17519 }, { "epoch": 2.2287240809057374, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.595917701721191, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8507718443870544, "num_tokens": 668493038.0, "step": 17520 }, { "epoch": 2.228851291184328, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.675286293029785, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8798599243164062, "num_tokens": 668528087.0, "step": 17521 }, { "epoch": 2.228978501462918, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.675725936889648, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8751040101051331, "num_tokens": 668562325.0, "step": 17522 }, { "epoch": 2.2291057117415085, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.616124153137207, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8752529621124268, "num_tokens": 668601336.0, "step": 17523 }, { "epoch": 2.229232922020099, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.631477355957031, "learning_rate": 1e-06, "loss": 0.3358, "mean_token_accuracy": 0.8914397954940796, "num_tokens": 668637113.0, "step": 17524 }, { "epoch": 2.2293601322986896, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.668556213378906, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8686285614967346, "num_tokens": 668675645.0, "step": 17525 }, { "epoch": 2.22948734257728, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.65737247467041, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8744727373123169, "num_tokens": 668706480.0, "step": 17526 }, { "epoch": 2.2296145528558706, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.662986755371094, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8675584197044373, "num_tokens": 668743956.0, "step": 17527 }, { "epoch": 2.229741763134461, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.66979694366455, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8818575143814087, "num_tokens": 668779134.0, "step": 17528 }, { "epoch": 2.2298689734130517, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.664278984069824, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8743849992752075, "num_tokens": 668814978.0, "step": 17529 }, { "epoch": 2.229996183691642, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.639925956726074, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8576244115829468, "num_tokens": 668854194.0, "step": 17530 }, { "epoch": 2.2301233939702327, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.67878532409668, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8650461435317993, "num_tokens": 668893095.0, "step": 17531 }, { "epoch": 2.2302506042488233, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.65560245513916, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8804059624671936, "num_tokens": 668930626.0, "step": 17532 }, { "epoch": 2.230377814527414, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.69582462310791, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8807414770126343, "num_tokens": 668960941.0, "step": 17533 }, { "epoch": 2.2305050248060043, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.606693267822266, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.876225471496582, "num_tokens": 668997771.0, "step": 17534 }, { "epoch": 2.230632235084595, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.603985786437988, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8860342502593994, "num_tokens": 669038714.0, "step": 17535 }, { "epoch": 2.2307594453631854, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.688082695007324, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.8814469575881958, "num_tokens": 669072336.0, "step": 17536 }, { "epoch": 2.230886655641776, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.6474609375, "learning_rate": 1e-06, "loss": 0.3652, "mean_token_accuracy": 0.8832384347915649, "num_tokens": 669110303.0, "step": 17537 }, { "epoch": 2.2310138659203664, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.644855499267578, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8632991313934326, "num_tokens": 669146874.0, "step": 17538 }, { "epoch": 2.231141076198957, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.666077613830566, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8859549164772034, "num_tokens": 669181682.0, "step": 17539 }, { "epoch": 2.2312682864775475, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.6383695602417, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.873092532157898, "num_tokens": 669227201.0, "step": 17540 }, { "epoch": 2.231395496756138, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.639327049255371, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8735758066177368, "num_tokens": 669265915.0, "step": 17541 }, { "epoch": 2.2315227070347285, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.6492338180542, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8775925636291504, "num_tokens": 669304130.0, "step": 17542 }, { "epoch": 2.231649917313319, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.624701499938965, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8715097904205322, "num_tokens": 669344225.0, "step": 17543 }, { "epoch": 2.2317771275919096, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.639486312866211, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8553236722946167, "num_tokens": 669385126.0, "step": 17544 }, { "epoch": 2.2319043378705, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.59883975982666, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8617355823516846, "num_tokens": 669425774.0, "step": 17545 }, { "epoch": 2.2320315481490907, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.703718185424805, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8638400435447693, "num_tokens": 669459789.0, "step": 17546 }, { "epoch": 2.2321587584276807, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.597823143005371, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8878989815711975, "num_tokens": 669498709.0, "step": 17547 }, { "epoch": 2.2322859687062713, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.620863914489746, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8781051635742188, "num_tokens": 669536021.0, "step": 17548 }, { "epoch": 2.232413178984862, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.664274215698242, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8619774580001831, "num_tokens": 669574316.0, "step": 17549 }, { "epoch": 2.2325403892634523, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.602574348449707, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8638368844985962, "num_tokens": 669611826.0, "step": 17550 }, { "epoch": 2.232667599542043, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.62346363067627, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8704506158828735, "num_tokens": 669655277.0, "step": 17551 }, { "epoch": 2.2327948098206334, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.606084823608398, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.866523265838623, "num_tokens": 669693481.0, "step": 17552 }, { "epoch": 2.232922020099224, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.708261489868164, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8723835945129395, "num_tokens": 669724104.0, "step": 17553 }, { "epoch": 2.2330492303778144, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.619917869567871, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8557342290878296, "num_tokens": 669768725.0, "step": 17554 }, { "epoch": 2.233176440656405, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.620078086853027, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8815137147903442, "num_tokens": 669809508.0, "step": 17555 }, { "epoch": 2.2333036509349955, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.646621704101562, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8742249011993408, "num_tokens": 669848047.0, "step": 17556 }, { "epoch": 2.233430861213586, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.633013725280762, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.8891923427581787, "num_tokens": 669881274.0, "step": 17557 }, { "epoch": 2.2335580714921766, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.68671989440918, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8672303557395935, "num_tokens": 669918517.0, "step": 17558 }, { "epoch": 2.233685281770767, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.63870620727539, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8649722933769226, "num_tokens": 669960990.0, "step": 17559 }, { "epoch": 2.2338124920493576, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.704778671264648, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8791000247001648, "num_tokens": 669995542.0, "step": 17560 }, { "epoch": 2.233939702327948, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.655180931091309, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.873774528503418, "num_tokens": 670036798.0, "step": 17561 }, { "epoch": 2.2340669126065387, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.675020217895508, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8719651699066162, "num_tokens": 670071763.0, "step": 17562 }, { "epoch": 2.234194122885129, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.69624137878418, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8728077411651611, "num_tokens": 670100864.0, "step": 17563 }, { "epoch": 2.2343213331637197, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.73573112487793, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8616437911987305, "num_tokens": 670144114.0, "step": 17564 }, { "epoch": 2.2344485434423103, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.585638046264648, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8766547441482544, "num_tokens": 670182719.0, "step": 17565 }, { "epoch": 2.234575753720901, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.674755096435547, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8768657445907593, "num_tokens": 670218059.0, "step": 17566 }, { "epoch": 2.2347029639994913, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.660078048706055, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8691422939300537, "num_tokens": 670259589.0, "step": 17567 }, { "epoch": 2.234830174278082, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.67675495147705, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8646520376205444, "num_tokens": 670301755.0, "step": 17568 }, { "epoch": 2.2349573845566724, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.648730278015137, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8774024248123169, "num_tokens": 670339205.0, "step": 17569 }, { "epoch": 2.235084594835263, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.683155059814453, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.875536322593689, "num_tokens": 670382961.0, "step": 17570 }, { "epoch": 2.2352118051138534, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.690042495727539, "learning_rate": 1e-06, "loss": 0.4389, "mean_token_accuracy": 0.8636058568954468, "num_tokens": 670420747.0, "step": 17571 }, { "epoch": 2.2353390153924435, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.674952507019043, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8897501230239868, "num_tokens": 670457906.0, "step": 17572 }, { "epoch": 2.235466225671034, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.609731674194336, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8614080548286438, "num_tokens": 670493187.0, "step": 17573 }, { "epoch": 2.2355934359496246, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.619468688964844, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8866822123527527, "num_tokens": 670529750.0, "step": 17574 }, { "epoch": 2.235720646228215, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.682157516479492, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8703857660293579, "num_tokens": 670570147.0, "step": 17575 }, { "epoch": 2.2358478565068056, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.692228317260742, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8753147721290588, "num_tokens": 670604113.0, "step": 17576 }, { "epoch": 2.235975066785396, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.671339988708496, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8718292713165283, "num_tokens": 670645451.0, "step": 17577 }, { "epoch": 2.2361022770639867, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.724024772644043, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.878488302230835, "num_tokens": 670684055.0, "step": 17578 }, { "epoch": 2.236229487342577, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.656535148620605, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8642394542694092, "num_tokens": 670724397.0, "step": 17579 }, { "epoch": 2.2363566976211677, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.698729515075684, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8836145401000977, "num_tokens": 670761282.0, "step": 17580 }, { "epoch": 2.2364839078997583, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.72386360168457, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8779149055480957, "num_tokens": 670800007.0, "step": 17581 }, { "epoch": 2.236611118178349, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.685246467590332, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8699784874916077, "num_tokens": 670840040.0, "step": 17582 }, { "epoch": 2.2367383284569393, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.682908058166504, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8661199808120728, "num_tokens": 670881193.0, "step": 17583 }, { "epoch": 2.23686553873553, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.70591926574707, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8760281801223755, "num_tokens": 670927275.0, "step": 17584 }, { "epoch": 2.2369927490141204, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.676658630371094, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8769047260284424, "num_tokens": 670967327.0, "step": 17585 }, { "epoch": 2.237119959292711, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.706083297729492, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8769035935401917, "num_tokens": 671009250.0, "step": 17586 }, { "epoch": 2.2372471695713014, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.629268646240234, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8774218559265137, "num_tokens": 671045068.0, "step": 17587 }, { "epoch": 2.237374379849892, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.694985389709473, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8604837656021118, "num_tokens": 671084786.0, "step": 17588 }, { "epoch": 2.2375015901284825, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.728450775146484, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8659625053405762, "num_tokens": 671123235.0, "step": 17589 }, { "epoch": 2.237628800407073, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.610209465026855, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8887894153594971, "num_tokens": 671161702.0, "step": 17590 }, { "epoch": 2.2377560106856635, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.66590404510498, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8843021988868713, "num_tokens": 671195487.0, "step": 17591 }, { "epoch": 2.237883220964254, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.63254165649414, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.87056964635849, "num_tokens": 671232288.0, "step": 17592 }, { "epoch": 2.2380104312428446, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.664764404296875, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8566603660583496, "num_tokens": 671275154.0, "step": 17593 }, { "epoch": 2.238137641521435, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.643381118774414, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8628261089324951, "num_tokens": 671312266.0, "step": 17594 }, { "epoch": 2.238264851800025, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.61428451538086, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.872952938079834, "num_tokens": 671349690.0, "step": 17595 }, { "epoch": 2.238392062078616, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.649856567382812, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8769503831863403, "num_tokens": 671387898.0, "step": 17596 }, { "epoch": 2.2385192723572063, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.65550708770752, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8520225286483765, "num_tokens": 671428054.0, "step": 17597 }, { "epoch": 2.238646482635797, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.657899856567383, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8793036341667175, "num_tokens": 671467096.0, "step": 17598 }, { "epoch": 2.2387736929143873, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.632183074951172, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8719706535339355, "num_tokens": 671506752.0, "step": 17599 }, { "epoch": 2.238900903192978, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.709413528442383, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8650712966918945, "num_tokens": 671542403.0, "step": 17600 }, { "epoch": 2.2390281134715684, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.67647647857666, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8844653964042664, "num_tokens": 671578115.0, "step": 17601 }, { "epoch": 2.239155323750159, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.602035522460938, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8837952017784119, "num_tokens": 671619854.0, "step": 17602 }, { "epoch": 2.2392825340287494, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.643720626831055, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.8866122364997864, "num_tokens": 671661617.0, "step": 17603 }, { "epoch": 2.23940974430734, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.625718116760254, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8627532124519348, "num_tokens": 671702231.0, "step": 17604 }, { "epoch": 2.2395369545859305, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.707494735717773, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.873305082321167, "num_tokens": 671741483.0, "step": 17605 }, { "epoch": 2.239664164864521, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.63260269165039, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8695285320281982, "num_tokens": 671780168.0, "step": 17606 }, { "epoch": 2.2397913751431116, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.696111679077148, "learning_rate": 1e-06, "loss": 0.4684, "mean_token_accuracy": 0.8507239818572998, "num_tokens": 671818980.0, "step": 17607 }, { "epoch": 2.239918585421702, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.669934272766113, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.873193621635437, "num_tokens": 671850010.0, "step": 17608 }, { "epoch": 2.2400457957002926, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.66781997680664, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8652832508087158, "num_tokens": 671888506.0, "step": 17609 }, { "epoch": 2.240173005978883, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.687637329101562, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8815341591835022, "num_tokens": 671930965.0, "step": 17610 }, { "epoch": 2.2403002162574737, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.649497985839844, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8822011947631836, "num_tokens": 671967472.0, "step": 17611 }, { "epoch": 2.240427426536064, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.723591804504395, "learning_rate": 1e-06, "loss": 0.3277, "mean_token_accuracy": 0.8944916725158691, "num_tokens": 672005986.0, "step": 17612 }, { "epoch": 2.2405546368146547, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.608879089355469, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8733469247817993, "num_tokens": 672044729.0, "step": 17613 }, { "epoch": 2.2406818470932452, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.65924072265625, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.874031662940979, "num_tokens": 672080141.0, "step": 17614 }, { "epoch": 2.2408090573718358, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.698169708251953, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.880715012550354, "num_tokens": 672119277.0, "step": 17615 }, { "epoch": 2.2409362676504263, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.65546703338623, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8714320063591003, "num_tokens": 672154205.0, "step": 17616 }, { "epoch": 2.241063477929017, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.63748836517334, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8724367618560791, "num_tokens": 672193667.0, "step": 17617 }, { "epoch": 2.2411906882076074, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.651839256286621, "learning_rate": 1e-06, "loss": 0.4953, "mean_token_accuracy": 0.8432133197784424, "num_tokens": 672232360.0, "step": 17618 }, { "epoch": 2.241317898486198, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.64981746673584, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8609590530395508, "num_tokens": 672264127.0, "step": 17619 }, { "epoch": 2.241445108764788, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.781781196594238, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8819005489349365, "num_tokens": 672300719.0, "step": 17620 }, { "epoch": 2.2415723190433785, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.633682250976562, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8838620781898499, "num_tokens": 672342796.0, "step": 17621 }, { "epoch": 2.241699529321969, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.567330360412598, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.874975323677063, "num_tokens": 672382597.0, "step": 17622 }, { "epoch": 2.2418267396005596, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.761781692504883, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8620132803916931, "num_tokens": 672427350.0, "step": 17623 }, { "epoch": 2.24195394987915, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.63558578491211, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8667356967926025, "num_tokens": 672468997.0, "step": 17624 }, { "epoch": 2.2420811601577406, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.716567039489746, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8791708946228027, "num_tokens": 672505061.0, "step": 17625 }, { "epoch": 2.242208370436331, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.654993057250977, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.871809720993042, "num_tokens": 672544241.0, "step": 17626 }, { "epoch": 2.2423355807149217, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.69713020324707, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8715860843658447, "num_tokens": 672586334.0, "step": 17627 }, { "epoch": 2.242462790993512, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.658242225646973, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8690328001976013, "num_tokens": 672628627.0, "step": 17628 }, { "epoch": 2.2425900012721027, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.68285846710205, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8635137677192688, "num_tokens": 672663046.0, "step": 17629 }, { "epoch": 2.2427172115506933, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.650395393371582, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8794207572937012, "num_tokens": 672703394.0, "step": 17630 }, { "epoch": 2.242844421829284, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.717878341674805, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8642213344573975, "num_tokens": 672743262.0, "step": 17631 }, { "epoch": 2.2429716321078743, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.62089729309082, "learning_rate": 1e-06, "loss": 0.4974, "mean_token_accuracy": 0.8400110006332397, "num_tokens": 672782662.0, "step": 17632 }, { "epoch": 2.243098842386465, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.65914535522461, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8705809712409973, "num_tokens": 672818400.0, "step": 17633 }, { "epoch": 2.2432260526650554, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.658875465393066, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8850034475326538, "num_tokens": 672858843.0, "step": 17634 }, { "epoch": 2.243353262943646, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.682568550109863, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8734039068222046, "num_tokens": 672900397.0, "step": 17635 }, { "epoch": 2.2434804732222364, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.673859596252441, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8707981109619141, "num_tokens": 672936025.0, "step": 17636 }, { "epoch": 2.243607683500827, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.637017250061035, "learning_rate": 1e-06, "loss": 0.4616, "mean_token_accuracy": 0.8544682264328003, "num_tokens": 672979180.0, "step": 17637 }, { "epoch": 2.2437348937794175, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.666484832763672, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8829546570777893, "num_tokens": 673018053.0, "step": 17638 }, { "epoch": 2.243862104058008, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.673510551452637, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8544242978096008, "num_tokens": 673057596.0, "step": 17639 }, { "epoch": 2.2439893143365985, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.673269271850586, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8847349882125854, "num_tokens": 673097740.0, "step": 17640 }, { "epoch": 2.244116524615189, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.636274337768555, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8784410953521729, "num_tokens": 673141821.0, "step": 17641 }, { "epoch": 2.2442437348937796, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.66735553741455, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8804033398628235, "num_tokens": 673184279.0, "step": 17642 }, { "epoch": 2.24437094517237, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.696596145629883, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8821969628334045, "num_tokens": 673224328.0, "step": 17643 }, { "epoch": 2.2444981554509607, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.706329345703125, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8598023653030396, "num_tokens": 673263328.0, "step": 17644 }, { "epoch": 2.2446253657295507, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.684741020202637, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8775001764297485, "num_tokens": 673301663.0, "step": 17645 }, { "epoch": 2.2447525760081413, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.69711971282959, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8701615333557129, "num_tokens": 673340606.0, "step": 17646 }, { "epoch": 2.244879786286732, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.715746879577637, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8704395890235901, "num_tokens": 673382201.0, "step": 17647 }, { "epoch": 2.2450069965653223, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.647819519042969, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8815357685089111, "num_tokens": 673423330.0, "step": 17648 }, { "epoch": 2.245134206843913, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.70805549621582, "learning_rate": 1e-06, "loss": 0.3407, "mean_token_accuracy": 0.888554573059082, "num_tokens": 673459690.0, "step": 17649 }, { "epoch": 2.2452614171225034, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.738966941833496, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8682582378387451, "num_tokens": 673494827.0, "step": 17650 }, { "epoch": 2.245388627401094, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.676470756530762, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8607052564620972, "num_tokens": 673540568.0, "step": 17651 }, { "epoch": 2.2455158376796844, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.68274211883545, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8877085447311401, "num_tokens": 673574022.0, "step": 17652 }, { "epoch": 2.245643047958275, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.699851036071777, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8693020939826965, "num_tokens": 673612493.0, "step": 17653 }, { "epoch": 2.2457702582368655, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.646292686462402, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8688369989395142, "num_tokens": 673653729.0, "step": 17654 }, { "epoch": 2.245897468515456, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.646366119384766, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8855669498443604, "num_tokens": 673695626.0, "step": 17655 }, { "epoch": 2.2460246787940465, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.643866539001465, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8801584839820862, "num_tokens": 673732188.0, "step": 17656 }, { "epoch": 2.246151889072637, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.660826683044434, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.865054726600647, "num_tokens": 673770259.0, "step": 17657 }, { "epoch": 2.2462790993512276, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.703825950622559, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8697289228439331, "num_tokens": 673805018.0, "step": 17658 }, { "epoch": 2.246406309629818, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.647591590881348, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8666342496871948, "num_tokens": 673846058.0, "step": 17659 }, { "epoch": 2.2465335199084087, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.71798324584961, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.859626054763794, "num_tokens": 673886437.0, "step": 17660 }, { "epoch": 2.246660730186999, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.699339866638184, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8765380382537842, "num_tokens": 673925701.0, "step": 17661 }, { "epoch": 2.2467879404655897, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.661483764648438, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8757065534591675, "num_tokens": 673969854.0, "step": 17662 }, { "epoch": 2.2469151507441802, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.69560718536377, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8746556043624878, "num_tokens": 674005048.0, "step": 17663 }, { "epoch": 2.2470423610227708, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.648107528686523, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8678773641586304, "num_tokens": 674046548.0, "step": 17664 }, { "epoch": 2.2471695713013613, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.663100242614746, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8677992820739746, "num_tokens": 674086905.0, "step": 17665 }, { "epoch": 2.247296781579952, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.665771484375, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.8825111985206604, "num_tokens": 674124589.0, "step": 17666 }, { "epoch": 2.2474239918585424, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.681025505065918, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8853553533554077, "num_tokens": 674165135.0, "step": 17667 }, { "epoch": 2.247551202137133, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.669299125671387, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8679133057594299, "num_tokens": 674207269.0, "step": 17668 }, { "epoch": 2.2476784124157234, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.70337963104248, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.8891468048095703, "num_tokens": 674240806.0, "step": 17669 }, { "epoch": 2.2478056226943135, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.722902297973633, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8640732169151306, "num_tokens": 674274697.0, "step": 17670 }, { "epoch": 2.247932832972904, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.690744400024414, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8634730577468872, "num_tokens": 674311362.0, "step": 17671 }, { "epoch": 2.2480600432514946, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.721214294433594, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8798459768295288, "num_tokens": 674352505.0, "step": 17672 }, { "epoch": 2.248187253530085, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.654160499572754, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8779601454734802, "num_tokens": 674383938.0, "step": 17673 }, { "epoch": 2.2483144638086756, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.746685028076172, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8696648478507996, "num_tokens": 674421665.0, "step": 17674 }, { "epoch": 2.248441674087266, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.665492057800293, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8750736713409424, "num_tokens": 674466065.0, "step": 17675 }, { "epoch": 2.2485688843658567, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.692912101745605, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8747104406356812, "num_tokens": 674506868.0, "step": 17676 }, { "epoch": 2.248696094644447, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.723881721496582, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8682533502578735, "num_tokens": 674545756.0, "step": 17677 }, { "epoch": 2.2488233049230377, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.71237850189209, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8679994940757751, "num_tokens": 674590680.0, "step": 17678 }, { "epoch": 2.2489505152016283, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.624817848205566, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8782056570053101, "num_tokens": 674631624.0, "step": 17679 }, { "epoch": 2.249077725480219, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.73227596282959, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8678807020187378, "num_tokens": 674667365.0, "step": 17680 }, { "epoch": 2.2492049357588093, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.680359840393066, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8720982074737549, "num_tokens": 674706587.0, "step": 17681 }, { "epoch": 2.2493321460374, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.684220314025879, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8711557388305664, "num_tokens": 674744480.0, "step": 17682 }, { "epoch": 2.2494593563159904, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.664278030395508, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8730990290641785, "num_tokens": 674783104.0, "step": 17683 }, { "epoch": 2.249586566594581, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.680346488952637, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8654943704605103, "num_tokens": 674823488.0, "step": 17684 }, { "epoch": 2.2497137768731714, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.659401893615723, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8783137798309326, "num_tokens": 674857322.0, "step": 17685 }, { "epoch": 2.249840987151762, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.691028594970703, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8643190860748291, "num_tokens": 674901566.0, "step": 17686 }, { "epoch": 2.2499681974303525, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.680007934570312, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8791669607162476, "num_tokens": 674942803.0, "step": 17687 }, { "epoch": 2.250095407708943, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.625431060791016, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8640803098678589, "num_tokens": 674983160.0, "step": 17688 }, { "epoch": 2.2502226179875335, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.707805633544922, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8744597434997559, "num_tokens": 675021243.0, "step": 17689 }, { "epoch": 2.250349828266124, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.591644287109375, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.869406521320343, "num_tokens": 675059435.0, "step": 17690 }, { "epoch": 2.2504770385447146, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.722129821777344, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8757106065750122, "num_tokens": 675091689.0, "step": 17691 }, { "epoch": 2.250604248823305, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.667564392089844, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8675696849822998, "num_tokens": 675130368.0, "step": 17692 }, { "epoch": 2.250731459101895, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.699456214904785, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8758796453475952, "num_tokens": 675171745.0, "step": 17693 }, { "epoch": 2.250858669380486, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.72767162322998, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8715359568595886, "num_tokens": 675206052.0, "step": 17694 }, { "epoch": 2.2509858796590763, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.654853820800781, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8682859539985657, "num_tokens": 675243231.0, "step": 17695 }, { "epoch": 2.251113089937667, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.790443420410156, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.866018533706665, "num_tokens": 675273582.0, "step": 17696 }, { "epoch": 2.2512403002162573, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.61905288696289, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8685446381568909, "num_tokens": 675314247.0, "step": 17697 }, { "epoch": 2.251367510494848, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.700813293457031, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8745702505111694, "num_tokens": 675355675.0, "step": 17698 }, { "epoch": 2.2514947207734384, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.70637035369873, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8710556626319885, "num_tokens": 675394290.0, "step": 17699 }, { "epoch": 2.251621931052029, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.718939781188965, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8760735392570496, "num_tokens": 675434251.0, "step": 17700 }, { "epoch": 2.2517491413306194, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.610221862792969, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.878822922706604, "num_tokens": 675474731.0, "step": 17701 }, { "epoch": 2.25187635160921, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.750517845153809, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.868495523929596, "num_tokens": 675515180.0, "step": 17702 }, { "epoch": 2.2520035618878005, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.682158470153809, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8622481822967529, "num_tokens": 675553429.0, "step": 17703 }, { "epoch": 2.252130772166391, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.679634094238281, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8763066530227661, "num_tokens": 675592919.0, "step": 17704 }, { "epoch": 2.2522579824449815, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.716729164123535, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8802642226219177, "num_tokens": 675627528.0, "step": 17705 }, { "epoch": 2.252385192723572, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.684922218322754, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.880576491355896, "num_tokens": 675663721.0, "step": 17706 }, { "epoch": 2.2525124030021626, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.708845138549805, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8675265312194824, "num_tokens": 675705187.0, "step": 17707 }, { "epoch": 2.252639613280753, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.693222045898438, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8818703889846802, "num_tokens": 675752713.0, "step": 17708 }, { "epoch": 2.2527668235593437, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.703070640563965, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8739661574363708, "num_tokens": 675791009.0, "step": 17709 }, { "epoch": 2.252894033837934, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.663748741149902, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8765064477920532, "num_tokens": 675828890.0, "step": 17710 }, { "epoch": 2.2530212441165247, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.727502822875977, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8743120431900024, "num_tokens": 675862855.0, "step": 17711 }, { "epoch": 2.2531484543951152, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.661766052246094, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8713568449020386, "num_tokens": 675903032.0, "step": 17712 }, { "epoch": 2.2532756646737058, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.670730590820312, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8690165281295776, "num_tokens": 675943035.0, "step": 17713 }, { "epoch": 2.2534028749522963, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.700342178344727, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8751903772354126, "num_tokens": 675983073.0, "step": 17714 }, { "epoch": 2.253530085230887, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.653508186340332, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8785130977630615, "num_tokens": 676022341.0, "step": 17715 }, { "epoch": 2.2536572955094774, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.662822723388672, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.869786262512207, "num_tokens": 676060907.0, "step": 17716 }, { "epoch": 2.253784505788068, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.693665504455566, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8715420365333557, "num_tokens": 676100693.0, "step": 17717 }, { "epoch": 2.253911716066658, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.71285343170166, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8859269618988037, "num_tokens": 676137526.0, "step": 17718 }, { "epoch": 2.254038926345249, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.72426700592041, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8756523728370667, "num_tokens": 676177245.0, "step": 17719 }, { "epoch": 2.254166136623839, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.683683395385742, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8795337677001953, "num_tokens": 676214568.0, "step": 17720 }, { "epoch": 2.2542933469024296, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.705487251281738, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.874148964881897, "num_tokens": 676252552.0, "step": 17721 }, { "epoch": 2.25442055718102, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.702808380126953, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.855202317237854, "num_tokens": 676290480.0, "step": 17722 }, { "epoch": 2.2545477674596106, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.641444206237793, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.885037899017334, "num_tokens": 676324950.0, "step": 17723 }, { "epoch": 2.254674977738201, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.675567626953125, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.864637017250061, "num_tokens": 676363851.0, "step": 17724 }, { "epoch": 2.2548021880167917, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.750631332397461, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8816118240356445, "num_tokens": 676399920.0, "step": 17725 }, { "epoch": 2.254929398295382, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.678338050842285, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8705358505249023, "num_tokens": 676443243.0, "step": 17726 }, { "epoch": 2.2550566085739727, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.75403881072998, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8597882986068726, "num_tokens": 676480272.0, "step": 17727 }, { "epoch": 2.2551838188525632, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.719073295593262, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8688480257987976, "num_tokens": 676516630.0, "step": 17728 }, { "epoch": 2.2553110291311538, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.719005584716797, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8775805234909058, "num_tokens": 676557842.0, "step": 17729 }, { "epoch": 2.2554382394097443, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.755826950073242, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8833222389221191, "num_tokens": 676599413.0, "step": 17730 }, { "epoch": 2.255565449688335, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.741257667541504, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8657710552215576, "num_tokens": 676637791.0, "step": 17731 }, { "epoch": 2.2556926599669254, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.708443641662598, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8694506883621216, "num_tokens": 676677108.0, "step": 17732 }, { "epoch": 2.255819870245516, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.691893577575684, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8746308088302612, "num_tokens": 676720698.0, "step": 17733 }, { "epoch": 2.2559470805241064, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.691960334777832, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8627467155456543, "num_tokens": 676763719.0, "step": 17734 }, { "epoch": 2.256074290802697, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.663276672363281, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8741710186004639, "num_tokens": 676801083.0, "step": 17735 }, { "epoch": 2.2562015010812875, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.703227996826172, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8658863306045532, "num_tokens": 676840861.0, "step": 17736 }, { "epoch": 2.256328711359878, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.73471736907959, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8855841159820557, "num_tokens": 676876864.0, "step": 17737 }, { "epoch": 2.2564559216384685, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.662160873413086, "learning_rate": 1e-06, "loss": 0.3387, "mean_token_accuracy": 0.8920316696166992, "num_tokens": 676920983.0, "step": 17738 }, { "epoch": 2.256583131917059, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.743160247802734, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8819360136985779, "num_tokens": 676961294.0, "step": 17739 }, { "epoch": 2.2567103421956496, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.720526695251465, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8659622669219971, "num_tokens": 676996318.0, "step": 17740 }, { "epoch": 2.2568375524742397, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.684317588806152, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8750245571136475, "num_tokens": 677034083.0, "step": 17741 }, { "epoch": 2.2569647627528306, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.762879371643066, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8741222620010376, "num_tokens": 677071206.0, "step": 17742 }, { "epoch": 2.2570919730314207, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.740006446838379, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8791114091873169, "num_tokens": 677103573.0, "step": 17743 }, { "epoch": 2.2572191833100113, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.684261322021484, "learning_rate": 1e-06, "loss": 0.3688, "mean_token_accuracy": 0.8827501535415649, "num_tokens": 677141406.0, "step": 17744 }, { "epoch": 2.257346393588602, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.71129322052002, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8627922534942627, "num_tokens": 677185968.0, "step": 17745 }, { "epoch": 2.2574736038671923, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.734244346618652, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8656827211380005, "num_tokens": 677229004.0, "step": 17746 }, { "epoch": 2.257600814145783, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.639504432678223, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8831113576889038, "num_tokens": 677266317.0, "step": 17747 }, { "epoch": 2.2577280244243734, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.77492904663086, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.882750391960144, "num_tokens": 677302344.0, "step": 17748 }, { "epoch": 2.257855234702964, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.73169994354248, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8637897968292236, "num_tokens": 677342103.0, "step": 17749 }, { "epoch": 2.2579824449815544, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.719012260437012, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.861557126045227, "num_tokens": 677378602.0, "step": 17750 }, { "epoch": 2.258109655260145, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.706940650939941, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8727938532829285, "num_tokens": 677415005.0, "step": 17751 }, { "epoch": 2.2582368655387355, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.747435569763184, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8556423187255859, "num_tokens": 677457908.0, "step": 17752 }, { "epoch": 2.258364075817326, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.763238906860352, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8635985255241394, "num_tokens": 677492759.0, "step": 17753 }, { "epoch": 2.2584912860959165, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.739089965820312, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8708381652832031, "num_tokens": 677535771.0, "step": 17754 }, { "epoch": 2.258618496374507, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.716717720031738, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.885942816734314, "num_tokens": 677575325.0, "step": 17755 }, { "epoch": 2.2587457066530976, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.75140380859375, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8738278150558472, "num_tokens": 677614132.0, "step": 17756 }, { "epoch": 2.258872916931688, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.73432731628418, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8677784204483032, "num_tokens": 677645657.0, "step": 17757 }, { "epoch": 2.2590001272102787, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.72498607635498, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8808243870735168, "num_tokens": 677686238.0, "step": 17758 }, { "epoch": 2.259127337488869, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.702781677246094, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8772890567779541, "num_tokens": 677724080.0, "step": 17759 }, { "epoch": 2.2592545477674597, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.667799949645996, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8731062412261963, "num_tokens": 677765768.0, "step": 17760 }, { "epoch": 2.2593817580460502, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.707432746887207, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8735854625701904, "num_tokens": 677803937.0, "step": 17761 }, { "epoch": 2.2595089683246408, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.70347785949707, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.8829619288444519, "num_tokens": 677839685.0, "step": 17762 }, { "epoch": 2.2596361786032313, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.765889167785645, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8580383658409119, "num_tokens": 677880411.0, "step": 17763 }, { "epoch": 2.259763388881822, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.716768264770508, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8810683488845825, "num_tokens": 677921177.0, "step": 17764 }, { "epoch": 2.2598905991604123, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.693947792053223, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.883472204208374, "num_tokens": 677956393.0, "step": 17765 }, { "epoch": 2.2600178094390024, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.715400695800781, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8707317113876343, "num_tokens": 677988364.0, "step": 17766 }, { "epoch": 2.2601450197175934, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.7001314163208, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8814462423324585, "num_tokens": 678028093.0, "step": 17767 }, { "epoch": 2.2602722299961835, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.764496803283691, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8745123147964478, "num_tokens": 678068174.0, "step": 17768 }, { "epoch": 2.260399440274774, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.68391227722168, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8742648363113403, "num_tokens": 678105784.0, "step": 17769 }, { "epoch": 2.2605266505533645, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.737321853637695, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8845706582069397, "num_tokens": 678142873.0, "step": 17770 }, { "epoch": 2.260653860831955, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.778992652893066, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8700400590896606, "num_tokens": 678181527.0, "step": 17771 }, { "epoch": 2.2607810711105456, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.730673789978027, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.875835120677948, "num_tokens": 678221225.0, "step": 17772 }, { "epoch": 2.260908281389136, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.709831237792969, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8813537359237671, "num_tokens": 678254767.0, "step": 17773 }, { "epoch": 2.2610354916677267, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.795552253723145, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8761320114135742, "num_tokens": 678290830.0, "step": 17774 }, { "epoch": 2.261162701946317, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.693024635314941, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8840922117233276, "num_tokens": 678327102.0, "step": 17775 }, { "epoch": 2.2612899122249077, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.808215141296387, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8859864473342896, "num_tokens": 678364872.0, "step": 17776 }, { "epoch": 2.2614171225034982, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.782407760620117, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.867449939250946, "num_tokens": 678399233.0, "step": 17777 }, { "epoch": 2.2615443327820888, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.7236967086792, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8809000253677368, "num_tokens": 678434577.0, "step": 17778 }, { "epoch": 2.2616715430606793, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.771967887878418, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.879576563835144, "num_tokens": 678466835.0, "step": 17779 }, { "epoch": 2.26179875333927, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.773455619812012, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8881711363792419, "num_tokens": 678506817.0, "step": 17780 }, { "epoch": 2.2619259636178604, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 13.676765441894531, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8744328022003174, "num_tokens": 678543734.0, "step": 17781 }, { "epoch": 2.262053173896451, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.753483772277832, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8698175549507141, "num_tokens": 678585066.0, "step": 17782 }, { "epoch": 2.2621803841750414, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.7654390335083, "learning_rate": 1e-06, "loss": 0.346, "mean_token_accuracy": 0.888249933719635, "num_tokens": 678614951.0, "step": 17783 }, { "epoch": 2.262307594453632, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.692840576171875, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8760615587234497, "num_tokens": 678649671.0, "step": 17784 }, { "epoch": 2.2624348047322225, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.678993225097656, "learning_rate": 1e-06, "loss": 0.3458, "mean_token_accuracy": 0.8901311159133911, "num_tokens": 678693418.0, "step": 17785 }, { "epoch": 2.262562015010813, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.726690292358398, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.883992075920105, "num_tokens": 678730891.0, "step": 17786 }, { "epoch": 2.2626892252894035, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.70887279510498, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8826377391815186, "num_tokens": 678766369.0, "step": 17787 }, { "epoch": 2.262816435567994, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.759759902954102, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8625284433364868, "num_tokens": 678809118.0, "step": 17788 }, { "epoch": 2.2629436458465846, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.726510047912598, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8723897933959961, "num_tokens": 678850057.0, "step": 17789 }, { "epoch": 2.263070856125175, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.7285737991333, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8723922967910767, "num_tokens": 678890253.0, "step": 17790 }, { "epoch": 2.263198066403765, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.731145858764648, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8885847330093384, "num_tokens": 678922535.0, "step": 17791 }, { "epoch": 2.263325276682356, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.746761322021484, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.872768759727478, "num_tokens": 678956136.0, "step": 17792 }, { "epoch": 2.2634524869609463, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.741853713989258, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8778153657913208, "num_tokens": 678990864.0, "step": 17793 }, { "epoch": 2.263579697239537, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.752286911010742, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8862193822860718, "num_tokens": 679028345.0, "step": 17794 }, { "epoch": 2.2637069075181273, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.724532127380371, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8619462847709656, "num_tokens": 679073679.0, "step": 17795 }, { "epoch": 2.263834117796718, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.735570907592773, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8725756406784058, "num_tokens": 679109274.0, "step": 17796 }, { "epoch": 2.2639613280753084, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.779777526855469, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8751318454742432, "num_tokens": 679141644.0, "step": 17797 }, { "epoch": 2.264088538353899, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.75455379486084, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8585934638977051, "num_tokens": 679184351.0, "step": 17798 }, { "epoch": 2.2642157486324894, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.701977729797363, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.870316743850708, "num_tokens": 679222509.0, "step": 17799 }, { "epoch": 2.26434295891108, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.765048027038574, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8579925298690796, "num_tokens": 679263362.0, "step": 17800 }, { "epoch": 2.2644701691896705, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.727217674255371, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8625388145446777, "num_tokens": 679307859.0, "step": 17801 }, { "epoch": 2.264597379468261, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.709432601928711, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8842934370040894, "num_tokens": 679344801.0, "step": 17802 }, { "epoch": 2.2647245897468515, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.732784271240234, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8664839863777161, "num_tokens": 679382795.0, "step": 17803 }, { "epoch": 2.264851800025442, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.643926620483398, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8718307018280029, "num_tokens": 679425282.0, "step": 17804 }, { "epoch": 2.2649790103040326, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.713277816772461, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8615601062774658, "num_tokens": 679466881.0, "step": 17805 }, { "epoch": 2.265106220582623, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.760639190673828, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8865484595298767, "num_tokens": 679509268.0, "step": 17806 }, { "epoch": 2.2652334308612136, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.670939445495605, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8712382912635803, "num_tokens": 679545082.0, "step": 17807 }, { "epoch": 2.265360641139804, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.760726928710938, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8758395314216614, "num_tokens": 679581795.0, "step": 17808 }, { "epoch": 2.2654878514183947, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.758613586425781, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8716357946395874, "num_tokens": 679618978.0, "step": 17809 }, { "epoch": 2.2656150616969852, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.738890647888184, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8767902851104736, "num_tokens": 679655002.0, "step": 17810 }, { "epoch": 2.2657422719755758, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.72276496887207, "learning_rate": 1e-06, "loss": 0.344, "mean_token_accuracy": 0.8881603479385376, "num_tokens": 679691143.0, "step": 17811 }, { "epoch": 2.2658694822541663, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.727524757385254, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8759769797325134, "num_tokens": 679729168.0, "step": 17812 }, { "epoch": 2.265996692532757, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.68478012084961, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.87296462059021, "num_tokens": 679773385.0, "step": 17813 }, { "epoch": 2.2661239028113473, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.745952606201172, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8854168653488159, "num_tokens": 679813884.0, "step": 17814 }, { "epoch": 2.266251113089938, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.66732120513916, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8754581212997437, "num_tokens": 679850834.0, "step": 17815 }, { "epoch": 2.266378323368528, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.74571704864502, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8801577091217041, "num_tokens": 679890210.0, "step": 17816 }, { "epoch": 2.266505533647119, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.708330154418945, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8824687004089355, "num_tokens": 679926252.0, "step": 17817 }, { "epoch": 2.266632743925709, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.740944862365723, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8709293603897095, "num_tokens": 679959626.0, "step": 17818 }, { "epoch": 2.2667599542042995, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.712311744689941, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8689378499984741, "num_tokens": 679999973.0, "step": 17819 }, { "epoch": 2.26688716448289, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.68487548828125, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8676950931549072, "num_tokens": 680041081.0, "step": 17820 }, { "epoch": 2.2670143747614806, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.686620712280273, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8789488077163696, "num_tokens": 680078953.0, "step": 17821 }, { "epoch": 2.267141585040071, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.722393989562988, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8828389644622803, "num_tokens": 680112309.0, "step": 17822 }, { "epoch": 2.2672687953186617, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.73716926574707, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8693629503250122, "num_tokens": 680148607.0, "step": 17823 }, { "epoch": 2.267396005597252, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.688138961791992, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.8912938833236694, "num_tokens": 680189617.0, "step": 17824 }, { "epoch": 2.2675232158758427, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.736296653747559, "learning_rate": 1e-06, "loss": 0.3459, "mean_token_accuracy": 0.8899115324020386, "num_tokens": 680227940.0, "step": 17825 }, { "epoch": 2.2676504261544332, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.778267860412598, "learning_rate": 1e-06, "loss": 0.3595, "mean_token_accuracy": 0.8867982625961304, "num_tokens": 680263473.0, "step": 17826 }, { "epoch": 2.2677776364330238, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.727424621582031, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8664566278457642, "num_tokens": 680297895.0, "step": 17827 }, { "epoch": 2.2679048467116143, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.747845649719238, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8763381242752075, "num_tokens": 680337658.0, "step": 17828 }, { "epoch": 2.268032056990205, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.704920768737793, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8678188323974609, "num_tokens": 680375912.0, "step": 17829 }, { "epoch": 2.2681592672687954, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.713908195495605, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8858404755592346, "num_tokens": 680415231.0, "step": 17830 }, { "epoch": 2.268286477547386, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.72280216217041, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8702057003974915, "num_tokens": 680460037.0, "step": 17831 }, { "epoch": 2.2684136878259764, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.689861297607422, "learning_rate": 1e-06, "loss": 0.4619, "mean_token_accuracy": 0.8520755767822266, "num_tokens": 680505521.0, "step": 17832 }, { "epoch": 2.268540898104567, "ewc_loss": 0.03564453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 13.72429084777832, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8876937627792358, "num_tokens": 680547682.0, "step": 17833 }, { "epoch": 2.2686681083831575, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.69058895111084, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8810499310493469, "num_tokens": 680587510.0, "step": 17834 }, { "epoch": 2.268795318661748, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.758835792541504, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8781366348266602, "num_tokens": 680624233.0, "step": 17835 }, { "epoch": 2.2689225289403385, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.688159942626953, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8700856566429138, "num_tokens": 680666091.0, "step": 17836 }, { "epoch": 2.269049739218929, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.719988822937012, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8808053731918335, "num_tokens": 680703062.0, "step": 17837 }, { "epoch": 2.2691769494975196, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.697591781616211, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.887589693069458, "num_tokens": 680735966.0, "step": 17838 }, { "epoch": 2.2693041597761097, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.769662857055664, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8746660351753235, "num_tokens": 680780010.0, "step": 17839 }, { "epoch": 2.2694313700547006, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.732177734375, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8736620545387268, "num_tokens": 680819516.0, "step": 17840 }, { "epoch": 2.2695585803332907, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.663495063781738, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.874275267124176, "num_tokens": 680861374.0, "step": 17841 }, { "epoch": 2.2696857906118812, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.749897956848145, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8664023280143738, "num_tokens": 680902447.0, "step": 17842 }, { "epoch": 2.2698130008904718, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.724347114562988, "learning_rate": 1e-06, "loss": 0.3404, "mean_token_accuracy": 0.8905463218688965, "num_tokens": 680942376.0, "step": 17843 }, { "epoch": 2.2699402111690623, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.746014595031738, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8751274347305298, "num_tokens": 680977060.0, "step": 17844 }, { "epoch": 2.270067421447653, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.739453315734863, "learning_rate": 1e-06, "loss": 0.4439, "mean_token_accuracy": 0.8599563241004944, "num_tokens": 681014504.0, "step": 17845 }, { "epoch": 2.2701946317262434, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.671055793762207, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8582493662834167, "num_tokens": 681053919.0, "step": 17846 }, { "epoch": 2.270321842004834, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.773160934448242, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8747194409370422, "num_tokens": 681090883.0, "step": 17847 }, { "epoch": 2.2704490522834244, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.752074241638184, "learning_rate": 1e-06, "loss": 0.3525, "mean_token_accuracy": 0.8876548409461975, "num_tokens": 681128819.0, "step": 17848 }, { "epoch": 2.270576262562015, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.70480728149414, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8653143644332886, "num_tokens": 681171434.0, "step": 17849 }, { "epoch": 2.2707034728406055, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.6995210647583, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8842650651931763, "num_tokens": 681206297.0, "step": 17850 }, { "epoch": 2.270830683119196, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.758538246154785, "learning_rate": 1e-06, "loss": 0.4297, "mean_token_accuracy": 0.864829421043396, "num_tokens": 681242521.0, "step": 17851 }, { "epoch": 2.2709578933977865, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.716592788696289, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.884577751159668, "num_tokens": 681282474.0, "step": 17852 }, { "epoch": 2.271085103676377, "ewc_loss": 0.035888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.734635353088379, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8708794116973877, "num_tokens": 681317280.0, "step": 17853 }, { "epoch": 2.2712123139549676, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.762056350708008, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8725658059120178, "num_tokens": 681349155.0, "step": 17854 }, { "epoch": 2.271339524233558, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.761454582214355, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8744949102401733, "num_tokens": 681380863.0, "step": 17855 }, { "epoch": 2.2714667345121486, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.751493453979492, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8786767721176147, "num_tokens": 681421970.0, "step": 17856 }, { "epoch": 2.271593944790739, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.750706672668457, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.88937908411026, "num_tokens": 681453511.0, "step": 17857 }, { "epoch": 2.2717211550693297, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.751626014709473, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8617830872535706, "num_tokens": 681496293.0, "step": 17858 }, { "epoch": 2.2718483653479202, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.724095344543457, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8677231073379517, "num_tokens": 681537074.0, "step": 17859 }, { "epoch": 2.2719755756265108, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.727255821228027, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8676823377609253, "num_tokens": 681578911.0, "step": 17860 }, { "epoch": 2.2721027859051013, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.714350700378418, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8713135123252869, "num_tokens": 681622134.0, "step": 17861 }, { "epoch": 2.272229996183692, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.722689628601074, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8753569722175598, "num_tokens": 681655919.0, "step": 17862 }, { "epoch": 2.2723572064622823, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.715633392333984, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8668811917304993, "num_tokens": 681696539.0, "step": 17863 }, { "epoch": 2.2724844167408724, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.738391876220703, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8712074160575867, "num_tokens": 681734630.0, "step": 17864 }, { "epoch": 2.2726116270194634, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.673800468444824, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8629080653190613, "num_tokens": 681779669.0, "step": 17865 }, { "epoch": 2.2727388372980535, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.704257011413574, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8719764947891235, "num_tokens": 681821920.0, "step": 17866 }, { "epoch": 2.272866047576644, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.73310375213623, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8725526928901672, "num_tokens": 681859239.0, "step": 17867 }, { "epoch": 2.2729932578552345, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.689745903015137, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8618488907814026, "num_tokens": 681896569.0, "step": 17868 }, { "epoch": 2.273120468133825, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.711569786071777, "learning_rate": 1e-06, "loss": 0.348, "mean_token_accuracy": 0.8889440298080444, "num_tokens": 681932549.0, "step": 17869 }, { "epoch": 2.2732476784124156, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.711859703063965, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8698129057884216, "num_tokens": 681973981.0, "step": 17870 }, { "epoch": 2.273374888691006, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.76119613647461, "learning_rate": 1e-06, "loss": 0.4839, "mean_token_accuracy": 0.8520967960357666, "num_tokens": 682009512.0, "step": 17871 }, { "epoch": 2.2735020989695967, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.751673698425293, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.872045636177063, "num_tokens": 682048058.0, "step": 17872 }, { "epoch": 2.273629309248187, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.692829132080078, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.86646968126297, "num_tokens": 682084181.0, "step": 17873 }, { "epoch": 2.2737565195267777, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.706216812133789, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8937538266181946, "num_tokens": 682119613.0, "step": 17874 }, { "epoch": 2.2738837298053682, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.672558784484863, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8770071864128113, "num_tokens": 682159501.0, "step": 17875 }, { "epoch": 2.2740109400839588, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.704503059387207, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8822296261787415, "num_tokens": 682197323.0, "step": 17876 }, { "epoch": 2.2741381503625493, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.678252220153809, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8749960064888, "num_tokens": 682231150.0, "step": 17877 }, { "epoch": 2.27426536064114, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.806884765625, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8759151101112366, "num_tokens": 682269351.0, "step": 17878 }, { "epoch": 2.2743925709197303, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.723223686218262, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8844516277313232, "num_tokens": 682303633.0, "step": 17879 }, { "epoch": 2.274519781198321, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.760149002075195, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8729278445243835, "num_tokens": 682340890.0, "step": 17880 }, { "epoch": 2.2746469914769114, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.735677719116211, "learning_rate": 1e-06, "loss": 0.3489, "mean_token_accuracy": 0.8893599510192871, "num_tokens": 682374253.0, "step": 17881 }, { "epoch": 2.274774201755502, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.713810920715332, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8738669753074646, "num_tokens": 682412675.0, "step": 17882 }, { "epoch": 2.2749014120340925, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.77359676361084, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8856956958770752, "num_tokens": 682447864.0, "step": 17883 }, { "epoch": 2.275028622312683, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.655583381652832, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8747216463088989, "num_tokens": 682486445.0, "step": 17884 }, { "epoch": 2.2751558325912735, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.776591300964355, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8741381168365479, "num_tokens": 682518173.0, "step": 17885 }, { "epoch": 2.275283042869864, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.70446491241455, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8738923072814941, "num_tokens": 682558627.0, "step": 17886 }, { "epoch": 2.2754102531484546, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.761601448059082, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8788400888442993, "num_tokens": 682600821.0, "step": 17887 }, { "epoch": 2.275537463427045, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.672945976257324, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.8624542355537415, "num_tokens": 682643947.0, "step": 17888 }, { "epoch": 2.275664673705635, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.766024589538574, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8688941597938538, "num_tokens": 682683045.0, "step": 17889 }, { "epoch": 2.275791883984226, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.696860313415527, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8749678134918213, "num_tokens": 682721617.0, "step": 17890 }, { "epoch": 2.2759190942628162, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.723666191101074, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8678458333015442, "num_tokens": 682758339.0, "step": 17891 }, { "epoch": 2.2760463045414068, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.726801872253418, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8795915842056274, "num_tokens": 682800501.0, "step": 17892 }, { "epoch": 2.2761735148199973, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.711149215698242, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8755577206611633, "num_tokens": 682833231.0, "step": 17893 }, { "epoch": 2.276300725098588, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.711333274841309, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8792087435722351, "num_tokens": 682873945.0, "step": 17894 }, { "epoch": 2.2764279353771784, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.733320236206055, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8728728294372559, "num_tokens": 682918329.0, "step": 17895 }, { "epoch": 2.276555145655769, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.729512214660645, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8795745372772217, "num_tokens": 682956277.0, "step": 17896 }, { "epoch": 2.2766823559343594, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.813874244689941, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8782246708869934, "num_tokens": 682990648.0, "step": 17897 }, { "epoch": 2.27680956621295, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.673524856567383, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8887158632278442, "num_tokens": 683029046.0, "step": 17898 }, { "epoch": 2.2769367764915405, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.816947937011719, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8751078248023987, "num_tokens": 683064541.0, "step": 17899 }, { "epoch": 2.277063986770131, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.749588966369629, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8747434616088867, "num_tokens": 683105015.0, "step": 17900 }, { "epoch": 2.2771911970487215, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.652300834655762, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8679749369621277, "num_tokens": 683142938.0, "step": 17901 }, { "epoch": 2.277318407327312, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.739143371582031, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8867812156677246, "num_tokens": 683181593.0, "step": 17902 }, { "epoch": 2.2774456176059026, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.711054801940918, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8635187149047852, "num_tokens": 683218641.0, "step": 17903 }, { "epoch": 2.277572827884493, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.765084266662598, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8586438894271851, "num_tokens": 683261479.0, "step": 17904 }, { "epoch": 2.2777000381630836, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.706036567687988, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8492046594619751, "num_tokens": 683300602.0, "step": 17905 }, { "epoch": 2.277827248441674, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.734576225280762, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8730300068855286, "num_tokens": 683337012.0, "step": 17906 }, { "epoch": 2.2779544587202647, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.73056697845459, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8826392292976379, "num_tokens": 683368949.0, "step": 17907 }, { "epoch": 2.2780816689988552, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.644901275634766, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8718742728233337, "num_tokens": 683408741.0, "step": 17908 }, { "epoch": 2.2782088792774458, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.733293533325195, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8640973567962646, "num_tokens": 683447843.0, "step": 17909 }, { "epoch": 2.2783360895560363, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.657215118408203, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8712214231491089, "num_tokens": 683489695.0, "step": 17910 }, { "epoch": 2.278463299834627, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.720255851745605, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8717445731163025, "num_tokens": 683528381.0, "step": 17911 }, { "epoch": 2.2785905101132173, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.70753288269043, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8754896521568298, "num_tokens": 683567261.0, "step": 17912 }, { "epoch": 2.278717720391808, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.75394058227539, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8601938486099243, "num_tokens": 683601083.0, "step": 17913 }, { "epoch": 2.278844930670398, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.739127159118652, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8678892850875854, "num_tokens": 683643494.0, "step": 17914 }, { "epoch": 2.278972140948989, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.71314811706543, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8697391152381897, "num_tokens": 683684267.0, "step": 17915 }, { "epoch": 2.279099351227579, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.773775100708008, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8634254336357117, "num_tokens": 683723292.0, "step": 17916 }, { "epoch": 2.2792265615061695, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.726930618286133, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8843280076980591, "num_tokens": 683761378.0, "step": 17917 }, { "epoch": 2.27935377178476, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.789165496826172, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.866661012172699, "num_tokens": 683794838.0, "step": 17918 }, { "epoch": 2.2794809820633506, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.762862205505371, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8557778596878052, "num_tokens": 683832307.0, "step": 17919 }, { "epoch": 2.279608192341941, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.713862419128418, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8671623468399048, "num_tokens": 683871848.0, "step": 17920 }, { "epoch": 2.2797354026205316, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.743672370910645, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8775579333305359, "num_tokens": 683915086.0, "step": 17921 }, { "epoch": 2.279862612899122, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.753665924072266, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8670814037322998, "num_tokens": 683945411.0, "step": 17922 }, { "epoch": 2.2799898231777127, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.760706901550293, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.874032735824585, "num_tokens": 683979664.0, "step": 17923 }, { "epoch": 2.2801170334563032, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.665696144104004, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8643506169319153, "num_tokens": 684025526.0, "step": 17924 }, { "epoch": 2.2802442437348938, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.81296443939209, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8692647218704224, "num_tokens": 684056992.0, "step": 17925 }, { "epoch": 2.2803714540134843, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.697053909301758, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8779314756393433, "num_tokens": 684094515.0, "step": 17926 }, { "epoch": 2.280498664292075, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.743924140930176, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8808001279830933, "num_tokens": 684132294.0, "step": 17927 }, { "epoch": 2.2806258745706653, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.789470672607422, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8764714002609253, "num_tokens": 684164010.0, "step": 17928 }, { "epoch": 2.280753084849256, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.746702194213867, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.881301999092102, "num_tokens": 684202077.0, "step": 17929 }, { "epoch": 2.2808802951278464, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.781064987182617, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8475661277770996, "num_tokens": 684240799.0, "step": 17930 }, { "epoch": 2.281007505406437, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.769771575927734, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8711795210838318, "num_tokens": 684284080.0, "step": 17931 }, { "epoch": 2.2811347156850275, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.799476623535156, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8578155040740967, "num_tokens": 684317915.0, "step": 17932 }, { "epoch": 2.281261925963618, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.737993240356445, "learning_rate": 1e-06, "loss": 0.2855, "mean_token_accuracy": 0.9105315804481506, "num_tokens": 684354665.0, "step": 17933 }, { "epoch": 2.2813891362422085, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.732675552368164, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8727673292160034, "num_tokens": 684392866.0, "step": 17934 }, { "epoch": 2.281516346520799, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.772602081298828, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8778359889984131, "num_tokens": 684428422.0, "step": 17935 }, { "epoch": 2.2816435567993896, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.717916488647461, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.8608072996139526, "num_tokens": 684467129.0, "step": 17936 }, { "epoch": 2.2817707670779797, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.84716510772705, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8668169975280762, "num_tokens": 684500093.0, "step": 17937 }, { "epoch": 2.2818979773565706, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.703495025634766, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8644300699234009, "num_tokens": 684541422.0, "step": 17938 }, { "epoch": 2.2820251876351607, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.776103973388672, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8705507516860962, "num_tokens": 684579164.0, "step": 17939 }, { "epoch": 2.2821523979137512, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.72278118133545, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8582807779312134, "num_tokens": 684618617.0, "step": 17940 }, { "epoch": 2.2822796081923418, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.731269836425781, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8596423268318176, "num_tokens": 684655525.0, "step": 17941 }, { "epoch": 2.2824068184709323, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.754343032836914, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8729441165924072, "num_tokens": 684694525.0, "step": 17942 }, { "epoch": 2.282534028749523, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.687668800354004, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8741370439529419, "num_tokens": 684733247.0, "step": 17943 }, { "epoch": 2.2826612390281134, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.749520301818848, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8859462738037109, "num_tokens": 684763642.0, "step": 17944 }, { "epoch": 2.282788449306704, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.765207290649414, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8796645402908325, "num_tokens": 684806164.0, "step": 17945 }, { "epoch": 2.2829156595852944, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.725409507751465, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8713734745979309, "num_tokens": 684844193.0, "step": 17946 }, { "epoch": 2.283042869863885, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.710469245910645, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8786008954048157, "num_tokens": 684882103.0, "step": 17947 }, { "epoch": 2.2831700801424755, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.764311790466309, "learning_rate": 1e-06, "loss": 0.3483, "mean_token_accuracy": 0.8874079585075378, "num_tokens": 684917667.0, "step": 17948 }, { "epoch": 2.283297290421066, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.779959678649902, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8777241706848145, "num_tokens": 684954007.0, "step": 17949 }, { "epoch": 2.2834245006996565, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.737736701965332, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8695871233940125, "num_tokens": 684994693.0, "step": 17950 }, { "epoch": 2.283551710978247, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.739716529846191, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8612198829650879, "num_tokens": 685032713.0, "step": 17951 }, { "epoch": 2.2836789212568376, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.80734634399414, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8676344156265259, "num_tokens": 685066597.0, "step": 17952 }, { "epoch": 2.283806131535428, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.719148635864258, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.86805260181427, "num_tokens": 685106243.0, "step": 17953 }, { "epoch": 2.2839333418140186, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.7850980758667, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8651617765426636, "num_tokens": 685145177.0, "step": 17954 }, { "epoch": 2.284060552092609, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.781962394714355, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.879226803779602, "num_tokens": 685189130.0, "step": 17955 }, { "epoch": 2.2841877623711997, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.778742790222168, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8778204917907715, "num_tokens": 685227536.0, "step": 17956 }, { "epoch": 2.28431497264979, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.766435623168945, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8787835836410522, "num_tokens": 685267213.0, "step": 17957 }, { "epoch": 2.2844421829283807, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.842764854431152, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.871210515499115, "num_tokens": 685301115.0, "step": 17958 }, { "epoch": 2.2845693932069713, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.704429626464844, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8786224126815796, "num_tokens": 685338497.0, "step": 17959 }, { "epoch": 2.284696603485562, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.781059265136719, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8633393049240112, "num_tokens": 685374983.0, "step": 17960 }, { "epoch": 2.2848238137641523, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.754156112670898, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8790467977523804, "num_tokens": 685408859.0, "step": 17961 }, { "epoch": 2.2849510240427424, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.756406784057617, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8697667717933655, "num_tokens": 685445103.0, "step": 17962 }, { "epoch": 2.2850782343213334, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.65694522857666, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8692913055419922, "num_tokens": 685484657.0, "step": 17963 }, { "epoch": 2.2852054445999235, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.701387405395508, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8817006945610046, "num_tokens": 685524034.0, "step": 17964 }, { "epoch": 2.285332654878514, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.706026077270508, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8791638612747192, "num_tokens": 685560368.0, "step": 17965 }, { "epoch": 2.2854598651571045, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.77630615234375, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8697870969772339, "num_tokens": 685598192.0, "step": 17966 }, { "epoch": 2.285587075435695, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.711861610412598, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.875925600528717, "num_tokens": 685636300.0, "step": 17967 }, { "epoch": 2.2857142857142856, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.779718399047852, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8711148500442505, "num_tokens": 685671457.0, "step": 17968 }, { "epoch": 2.285841495992876, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.722707748413086, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8629088401794434, "num_tokens": 685709687.0, "step": 17969 }, { "epoch": 2.2859687062714666, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.763819694519043, "learning_rate": 1e-06, "loss": 0.4704, "mean_token_accuracy": 0.8517417311668396, "num_tokens": 685747854.0, "step": 17970 }, { "epoch": 2.286095916550057, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.707244873046875, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8928771018981934, "num_tokens": 685785215.0, "step": 17971 }, { "epoch": 2.2862231268286477, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.805329322814941, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8776165843009949, "num_tokens": 685818871.0, "step": 17972 }, { "epoch": 2.2863503371072382, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.76429271697998, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8668851852416992, "num_tokens": 685859089.0, "step": 17973 }, { "epoch": 2.2864775473858288, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.737824440002441, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8644557595252991, "num_tokens": 685900346.0, "step": 17974 }, { "epoch": 2.2866047576644193, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.776350975036621, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8719112277030945, "num_tokens": 685938186.0, "step": 17975 }, { "epoch": 2.28673196794301, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.72396469116211, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8798866271972656, "num_tokens": 685974911.0, "step": 17976 }, { "epoch": 2.2868591782216003, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.743321418762207, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8753266334533691, "num_tokens": 686010430.0, "step": 17977 }, { "epoch": 2.286986388500191, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.810151100158691, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.8862364888191223, "num_tokens": 686042727.0, "step": 17978 }, { "epoch": 2.2871135987787814, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.783271789550781, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.875338613986969, "num_tokens": 686078819.0, "step": 17979 }, { "epoch": 2.287240809057372, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.748642921447754, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.875451922416687, "num_tokens": 686117760.0, "step": 17980 }, { "epoch": 2.2873680193359625, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.77804946899414, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8639254570007324, "num_tokens": 686152992.0, "step": 17981 }, { "epoch": 2.287495229614553, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.771248817443848, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8677393198013306, "num_tokens": 686195353.0, "step": 17982 }, { "epoch": 2.2876224398931435, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.73114013671875, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8685300350189209, "num_tokens": 686232990.0, "step": 17983 }, { "epoch": 2.287749650171734, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.822493553161621, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8774850964546204, "num_tokens": 686265933.0, "step": 17984 }, { "epoch": 2.2878768604503246, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.736745834350586, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8793692588806152, "num_tokens": 686307111.0, "step": 17985 }, { "epoch": 2.288004070728915, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.790169715881348, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8757242560386658, "num_tokens": 686343363.0, "step": 17986 }, { "epoch": 2.288131281007505, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.702451705932617, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8718322515487671, "num_tokens": 686380888.0, "step": 17987 }, { "epoch": 2.288258491286096, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.73982048034668, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8650737404823303, "num_tokens": 686417350.0, "step": 17988 }, { "epoch": 2.2883857015646862, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.733964920043945, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.881088376045227, "num_tokens": 686458658.0, "step": 17989 }, { "epoch": 2.2885129118432768, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.795370101928711, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8761368989944458, "num_tokens": 686493945.0, "step": 17990 }, { "epoch": 2.2886401221218673, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.739898681640625, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8731564283370972, "num_tokens": 686527608.0, "step": 17991 }, { "epoch": 2.288767332400458, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.782747268676758, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8830952048301697, "num_tokens": 686562272.0, "step": 17992 }, { "epoch": 2.2888945426790483, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.72497844696045, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8853272199630737, "num_tokens": 686599937.0, "step": 17993 }, { "epoch": 2.289021752957639, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.808884620666504, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.87386554479599, "num_tokens": 686632103.0, "step": 17994 }, { "epoch": 2.2891489632362294, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.725866317749023, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8787117004394531, "num_tokens": 686662802.0, "step": 17995 }, { "epoch": 2.28927617351482, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.741240501403809, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.8941401839256287, "num_tokens": 686699386.0, "step": 17996 }, { "epoch": 2.2894033837934105, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.758646965026855, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.876295804977417, "num_tokens": 686738369.0, "step": 17997 }, { "epoch": 2.289530594072001, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.761597633361816, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.866142749786377, "num_tokens": 686774543.0, "step": 17998 }, { "epoch": 2.2896578043505915, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.751551628112793, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8656355142593384, "num_tokens": 686807048.0, "step": 17999 }, { "epoch": 2.289785014629182, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.754217147827148, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8741617202758789, "num_tokens": 686844678.0, "step": 18000 }, { "epoch": 2.2899122249077726, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.808013916015625, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8733074069023132, "num_tokens": 686877721.0, "step": 18001 }, { "epoch": 2.290039435186363, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.727128028869629, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8668392300605774, "num_tokens": 686918329.0, "step": 18002 }, { "epoch": 2.2901666454649536, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.748889923095703, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8690152168273926, "num_tokens": 686960050.0, "step": 18003 }, { "epoch": 2.290293855743544, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.815226554870605, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8836243748664856, "num_tokens": 686997849.0, "step": 18004 }, { "epoch": 2.2904210660221347, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.75787353515625, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8769615888595581, "num_tokens": 687032363.0, "step": 18005 }, { "epoch": 2.290548276300725, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.818262100219727, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8582960367202759, "num_tokens": 687069606.0, "step": 18006 }, { "epoch": 2.2906754865793157, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.756704330444336, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8641996383666992, "num_tokens": 687112857.0, "step": 18007 }, { "epoch": 2.2908026968579063, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.756117820739746, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8717796206474304, "num_tokens": 687152245.0, "step": 18008 }, { "epoch": 2.290929907136497, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.773880004882812, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8733108639717102, "num_tokens": 687193988.0, "step": 18009 }, { "epoch": 2.2910571174150873, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.720401763916016, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.868743896484375, "num_tokens": 687232375.0, "step": 18010 }, { "epoch": 2.291184327693678, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.731241226196289, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.876900315284729, "num_tokens": 687274566.0, "step": 18011 }, { "epoch": 2.291311537972268, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.780550956726074, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8730746507644653, "num_tokens": 687313231.0, "step": 18012 }, { "epoch": 2.291438748250859, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.767505645751953, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8781853914260864, "num_tokens": 687354502.0, "step": 18013 }, { "epoch": 2.291565958529449, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.713739395141602, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8700079917907715, "num_tokens": 687399160.0, "step": 18014 }, { "epoch": 2.2916931688080395, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.791390419006348, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8630783557891846, "num_tokens": 687438775.0, "step": 18015 }, { "epoch": 2.29182037908663, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.744473457336426, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.87714684009552, "num_tokens": 687471435.0, "step": 18016 }, { "epoch": 2.2919475893652206, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.762104988098145, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.87485671043396, "num_tokens": 687508657.0, "step": 18017 }, { "epoch": 2.292074799643811, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.746966361999512, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8835200071334839, "num_tokens": 687545024.0, "step": 18018 }, { "epoch": 2.2922020099224016, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.779853820800781, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8779562711715698, "num_tokens": 687579432.0, "step": 18019 }, { "epoch": 2.292329220200992, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.798432350158691, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8637057542800903, "num_tokens": 687618378.0, "step": 18020 }, { "epoch": 2.2924564304795827, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.702886581420898, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8723840713500977, "num_tokens": 687657119.0, "step": 18021 }, { "epoch": 2.2925836407581732, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.8219575881958, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8628460764884949, "num_tokens": 687691924.0, "step": 18022 }, { "epoch": 2.2927108510367638, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.742238998413086, "learning_rate": 1e-06, "loss": 0.4678, "mean_token_accuracy": 0.8507950305938721, "num_tokens": 687724221.0, "step": 18023 }, { "epoch": 2.2928380613153543, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.815741539001465, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8753578066825867, "num_tokens": 687765049.0, "step": 18024 }, { "epoch": 2.292965271593945, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.719265937805176, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8640394806861877, "num_tokens": 687810222.0, "step": 18025 }, { "epoch": 2.2930924818725353, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.740255355834961, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.868802547454834, "num_tokens": 687851961.0, "step": 18026 }, { "epoch": 2.293219692151126, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.741637229919434, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8623154163360596, "num_tokens": 687890448.0, "step": 18027 }, { "epoch": 2.2933469024297164, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.776167869567871, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8817923665046692, "num_tokens": 687929509.0, "step": 18028 }, { "epoch": 2.293474112708307, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.741652488708496, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8633284568786621, "num_tokens": 687966814.0, "step": 18029 }, { "epoch": 2.2936013229868975, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.809170722961426, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8800578117370605, "num_tokens": 688006398.0, "step": 18030 }, { "epoch": 2.293728533265488, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.757756233215332, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8712868094444275, "num_tokens": 688044604.0, "step": 18031 }, { "epoch": 2.2938557435440785, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.767348289489746, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8682448863983154, "num_tokens": 688082969.0, "step": 18032 }, { "epoch": 2.293982953822669, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.827340126037598, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8796802759170532, "num_tokens": 688121442.0, "step": 18033 }, { "epoch": 2.2941101641012596, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.776436805725098, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.874030351638794, "num_tokens": 688161096.0, "step": 18034 }, { "epoch": 2.2942373743798496, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.75356674194336, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8735690116882324, "num_tokens": 688200782.0, "step": 18035 }, { "epoch": 2.2943645846584406, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.729011535644531, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8799535036087036, "num_tokens": 688241333.0, "step": 18036 }, { "epoch": 2.2944917949370307, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.82048225402832, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8709994554519653, "num_tokens": 688278297.0, "step": 18037 }, { "epoch": 2.2946190052156212, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.76763916015625, "learning_rate": 1e-06, "loss": 0.4768, "mean_token_accuracy": 0.8493476510047913, "num_tokens": 688318835.0, "step": 18038 }, { "epoch": 2.2947462154942118, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 13.739953994750977, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8749303817749023, "num_tokens": 688354814.0, "step": 18039 }, { "epoch": 2.2948734257728023, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.700896263122559, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.875497579574585, "num_tokens": 688390008.0, "step": 18040 }, { "epoch": 2.295000636051393, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.784696578979492, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8757642507553101, "num_tokens": 688430938.0, "step": 18041 }, { "epoch": 2.2951278463299833, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.755496978759766, "learning_rate": 1e-06, "loss": 0.5157, "mean_token_accuracy": 0.8428111672401428, "num_tokens": 688466730.0, "step": 18042 }, { "epoch": 2.295255056608574, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.799908638000488, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8721356391906738, "num_tokens": 688506666.0, "step": 18043 }, { "epoch": 2.2953822668871644, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.758415222167969, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8592573404312134, "num_tokens": 688547726.0, "step": 18044 }, { "epoch": 2.295509477165755, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.768641471862793, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8693663477897644, "num_tokens": 688587487.0, "step": 18045 }, { "epoch": 2.2956366874443455, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.772578239440918, "learning_rate": 1e-06, "loss": 0.423, "mean_token_accuracy": 0.8661859631538391, "num_tokens": 688632454.0, "step": 18046 }, { "epoch": 2.295763897722936, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.770282745361328, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8799649477005005, "num_tokens": 688668489.0, "step": 18047 }, { "epoch": 2.2958911080015265, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.780632972717285, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.876004695892334, "num_tokens": 688707899.0, "step": 18048 }, { "epoch": 2.296018318280117, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.767043113708496, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8779796361923218, "num_tokens": 688747171.0, "step": 18049 }, { "epoch": 2.2961455285587076, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.795557975769043, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8706872463226318, "num_tokens": 688789553.0, "step": 18050 }, { "epoch": 2.296272738837298, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.770395278930664, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8774045705795288, "num_tokens": 688827632.0, "step": 18051 }, { "epoch": 2.2963999491158886, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.779624938964844, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8810513019561768, "num_tokens": 688861505.0, "step": 18052 }, { "epoch": 2.296527159394479, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.814146041870117, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8669761419296265, "num_tokens": 688898270.0, "step": 18053 }, { "epoch": 2.2966543696730697, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.744837760925293, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8727885484695435, "num_tokens": 688937420.0, "step": 18054 }, { "epoch": 2.29678157995166, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.790040016174316, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8790576457977295, "num_tokens": 688972426.0, "step": 18055 }, { "epoch": 2.2969087902302507, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.897404670715332, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8725138306617737, "num_tokens": 689005801.0, "step": 18056 }, { "epoch": 2.2970360005088413, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.812774658203125, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8683146834373474, "num_tokens": 689039999.0, "step": 18057 }, { "epoch": 2.297163210787432, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.792587280273438, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8581235408782959, "num_tokens": 689079404.0, "step": 18058 }, { "epoch": 2.2972904210660223, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.819214820861816, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8504902124404907, "num_tokens": 689116197.0, "step": 18059 }, { "epoch": 2.2974176313446124, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.769238471984863, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.873828113079071, "num_tokens": 689153446.0, "step": 18060 }, { "epoch": 2.2975448416232034, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.874476432800293, "learning_rate": 1e-06, "loss": 0.4501, "mean_token_accuracy": 0.8559068441390991, "num_tokens": 689185310.0, "step": 18061 }, { "epoch": 2.2976720519017935, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.724559783935547, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8840181827545166, "num_tokens": 689224150.0, "step": 18062 }, { "epoch": 2.297799262180384, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.788350105285645, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8630560636520386, "num_tokens": 689261206.0, "step": 18063 }, { "epoch": 2.2979264724589745, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.852945327758789, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8777511119842529, "num_tokens": 689299129.0, "step": 18064 }, { "epoch": 2.298053682737565, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.732442855834961, "learning_rate": 1e-06, "loss": 0.3511, "mean_token_accuracy": 0.8883284330368042, "num_tokens": 689330554.0, "step": 18065 }, { "epoch": 2.2981808930161556, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.80319881439209, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8791936039924622, "num_tokens": 689364084.0, "step": 18066 }, { "epoch": 2.298308103294746, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.734786987304688, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8613722920417786, "num_tokens": 689405254.0, "step": 18067 }, { "epoch": 2.2984353135733366, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.819607734680176, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8633854389190674, "num_tokens": 689441109.0, "step": 18068 }, { "epoch": 2.298562523851927, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.82437515258789, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8734164834022522, "num_tokens": 689474937.0, "step": 18069 }, { "epoch": 2.2986897341305177, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.753719329833984, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8678184151649475, "num_tokens": 689515971.0, "step": 18070 }, { "epoch": 2.298816944409108, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.771939277648926, "learning_rate": 1e-06, "loss": 0.4668, "mean_token_accuracy": 0.851052463054657, "num_tokens": 689553620.0, "step": 18071 }, { "epoch": 2.2989441546876987, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.77225112915039, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8770750761032104, "num_tokens": 689592157.0, "step": 18072 }, { "epoch": 2.2990713649662893, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.773492813110352, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8864666223526001, "num_tokens": 689626889.0, "step": 18073 }, { "epoch": 2.29919857524488, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.790966033935547, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8846640586853027, "num_tokens": 689666161.0, "step": 18074 }, { "epoch": 2.2993257855234703, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.77303695678711, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8646202683448792, "num_tokens": 689713713.0, "step": 18075 }, { "epoch": 2.299452995802061, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.763946533203125, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8654912710189819, "num_tokens": 689755146.0, "step": 18076 }, { "epoch": 2.2995802060806514, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.75589370727539, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8810347318649292, "num_tokens": 689795925.0, "step": 18077 }, { "epoch": 2.299707416359242, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.83866024017334, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8834683299064636, "num_tokens": 689830673.0, "step": 18078 }, { "epoch": 2.2998346266378324, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.760775566101074, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8728929758071899, "num_tokens": 689863649.0, "step": 18079 }, { "epoch": 2.299961836916423, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.81227970123291, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8719110488891602, "num_tokens": 689901946.0, "step": 18080 }, { "epoch": 2.3000890471950135, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.74720573425293, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.866786003112793, "num_tokens": 689939968.0, "step": 18081 }, { "epoch": 2.300216257473604, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.693828582763672, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8817728757858276, "num_tokens": 689979792.0, "step": 18082 }, { "epoch": 2.3003434677521946, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.800697326660156, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8847265839576721, "num_tokens": 690013974.0, "step": 18083 }, { "epoch": 2.300470678030785, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.684582710266113, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8807564377784729, "num_tokens": 690056540.0, "step": 18084 }, { "epoch": 2.300597888309375, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.7352294921875, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8668869137763977, "num_tokens": 690101032.0, "step": 18085 }, { "epoch": 2.300725098587966, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.765116691589355, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8853894472122192, "num_tokens": 690135417.0, "step": 18086 }, { "epoch": 2.3008523088665562, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.717880249023438, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8748284578323364, "num_tokens": 690178552.0, "step": 18087 }, { "epoch": 2.3009795191451468, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.760376930236816, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8698838949203491, "num_tokens": 690221928.0, "step": 18088 }, { "epoch": 2.3011067294237373, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.787115097045898, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8791038990020752, "num_tokens": 690257040.0, "step": 18089 }, { "epoch": 2.301233939702328, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.82361125946045, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8732823133468628, "num_tokens": 690302720.0, "step": 18090 }, { "epoch": 2.3013611499809183, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.767908096313477, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.868965208530426, "num_tokens": 690337042.0, "step": 18091 }, { "epoch": 2.301488360259509, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.786829948425293, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8791197538375854, "num_tokens": 690380247.0, "step": 18092 }, { "epoch": 2.3016155705380994, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.755260467529297, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8768032789230347, "num_tokens": 690425361.0, "step": 18093 }, { "epoch": 2.30174278081669, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.735345840454102, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8676712512969971, "num_tokens": 690465137.0, "step": 18094 }, { "epoch": 2.3018699910952805, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.781835556030273, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.870479941368103, "num_tokens": 690501437.0, "step": 18095 }, { "epoch": 2.301997201373871, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.726183891296387, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8682011961936951, "num_tokens": 690544134.0, "step": 18096 }, { "epoch": 2.3021244116524615, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.781237602233887, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.881761908531189, "num_tokens": 690581062.0, "step": 18097 }, { "epoch": 2.302251621931052, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.762797355651855, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8777700066566467, "num_tokens": 690615145.0, "step": 18098 }, { "epoch": 2.3023788322096426, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.74403190612793, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8793737292289734, "num_tokens": 690655483.0, "step": 18099 }, { "epoch": 2.302506042488233, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.708849906921387, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8702307343482971, "num_tokens": 690694133.0, "step": 18100 }, { "epoch": 2.3026332527668236, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.795592308044434, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8513167500495911, "num_tokens": 690729939.0, "step": 18101 }, { "epoch": 2.302760463045414, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.770081520080566, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8790073990821838, "num_tokens": 690770139.0, "step": 18102 }, { "epoch": 2.3028876733240047, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.778569221496582, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8746200203895569, "num_tokens": 690811379.0, "step": 18103 }, { "epoch": 2.303014883602595, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.786097526550293, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8685867190361023, "num_tokens": 690844194.0, "step": 18104 }, { "epoch": 2.3031420938811857, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.76712703704834, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8662660121917725, "num_tokens": 690882129.0, "step": 18105 }, { "epoch": 2.3032693041597763, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.754319190979004, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8861606121063232, "num_tokens": 690915568.0, "step": 18106 }, { "epoch": 2.303396514438367, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.810822486877441, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8709669709205627, "num_tokens": 690953412.0, "step": 18107 }, { "epoch": 2.3035237247169573, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.738266944885254, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8697445392608643, "num_tokens": 690990047.0, "step": 18108 }, { "epoch": 2.303650934995548, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.78670597076416, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8778315186500549, "num_tokens": 691031010.0, "step": 18109 }, { "epoch": 2.303778145274138, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.796954154968262, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8743550777435303, "num_tokens": 691065830.0, "step": 18110 }, { "epoch": 2.303905355552729, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.786589622497559, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8842018842697144, "num_tokens": 691101418.0, "step": 18111 }, { "epoch": 2.304032565831319, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.779072761535645, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8592277765274048, "num_tokens": 691139447.0, "step": 18112 }, { "epoch": 2.3041597761099095, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.808643341064453, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8541467189788818, "num_tokens": 691168809.0, "step": 18113 }, { "epoch": 2.3042869863885, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.750534057617188, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8798648118972778, "num_tokens": 691200420.0, "step": 18114 }, { "epoch": 2.3044141966670906, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.775733947753906, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8673734068870544, "num_tokens": 691238587.0, "step": 18115 }, { "epoch": 2.304541406945681, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.80391788482666, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8711615204811096, "num_tokens": 691277137.0, "step": 18116 }, { "epoch": 2.3046686172242716, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.737330436706543, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.871356725692749, "num_tokens": 691319564.0, "step": 18117 }, { "epoch": 2.304795827502862, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.829455375671387, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8712762594223022, "num_tokens": 691359652.0, "step": 18118 }, { "epoch": 2.3049230377814527, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.785898208618164, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8722877502441406, "num_tokens": 691399705.0, "step": 18119 }, { "epoch": 2.305050248060043, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.70202350616455, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8653604984283447, "num_tokens": 691439447.0, "step": 18120 }, { "epoch": 2.3051774583386337, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.867783546447754, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8616689443588257, "num_tokens": 691477453.0, "step": 18121 }, { "epoch": 2.3053046686172243, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.714812278747559, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.872715950012207, "num_tokens": 691517766.0, "step": 18122 }, { "epoch": 2.305431878895815, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.872844696044922, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8718756437301636, "num_tokens": 691553956.0, "step": 18123 }, { "epoch": 2.3055590891744053, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.746673583984375, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8710546493530273, "num_tokens": 691592522.0, "step": 18124 }, { "epoch": 2.305686299452996, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.789316177368164, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8715643882751465, "num_tokens": 691629202.0, "step": 18125 }, { "epoch": 2.3058135097315864, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.763833045959473, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8756693005561829, "num_tokens": 691665324.0, "step": 18126 }, { "epoch": 2.305940720010177, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.767030715942383, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8823628425598145, "num_tokens": 691704223.0, "step": 18127 }, { "epoch": 2.3060679302887674, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.72881031036377, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8696596622467041, "num_tokens": 691745061.0, "step": 18128 }, { "epoch": 2.306195140567358, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.716487884521484, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8845276236534119, "num_tokens": 691783630.0, "step": 18129 }, { "epoch": 2.3063223508459485, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.763219833374023, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8715770840644836, "num_tokens": 691818283.0, "step": 18130 }, { "epoch": 2.306449561124539, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.773780822753906, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8635205030441284, "num_tokens": 691854585.0, "step": 18131 }, { "epoch": 2.3065767714031296, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.773038864135742, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.855429470539093, "num_tokens": 691891784.0, "step": 18132 }, { "epoch": 2.3067039816817196, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.795961380004883, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8742705583572388, "num_tokens": 691927389.0, "step": 18133 }, { "epoch": 2.3068311919603106, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.777095794677734, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8704520463943481, "num_tokens": 691962928.0, "step": 18134 }, { "epoch": 2.3069584022389007, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.700868606567383, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8719168901443481, "num_tokens": 692007064.0, "step": 18135 }, { "epoch": 2.3070856125174912, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.754095077514648, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8773285150527954, "num_tokens": 692040668.0, "step": 18136 }, { "epoch": 2.3072128227960818, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.792466163635254, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8733628988265991, "num_tokens": 692077710.0, "step": 18137 }, { "epoch": 2.3073400330746723, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.762681007385254, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8799899816513062, "num_tokens": 692112316.0, "step": 18138 }, { "epoch": 2.307467243353263, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.804274559020996, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8782284259796143, "num_tokens": 692153201.0, "step": 18139 }, { "epoch": 2.3075944536318533, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.768942832946777, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8662697672843933, "num_tokens": 692192942.0, "step": 18140 }, { "epoch": 2.307721663910444, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.743989944458008, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8864957094192505, "num_tokens": 692235515.0, "step": 18141 }, { "epoch": 2.3078488741890344, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.798908233642578, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8635386228561401, "num_tokens": 692271338.0, "step": 18142 }, { "epoch": 2.307976084467625, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.745067596435547, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8538516759872437, "num_tokens": 692311469.0, "step": 18143 }, { "epoch": 2.3081032947462155, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.734928131103516, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8552974462509155, "num_tokens": 692352052.0, "step": 18144 }, { "epoch": 2.308230505024806, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.832687377929688, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.8570583462715149, "num_tokens": 692382978.0, "step": 18145 }, { "epoch": 2.3083577153033965, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.80597972869873, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8748602271080017, "num_tokens": 692415651.0, "step": 18146 }, { "epoch": 2.308484925581987, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.761425018310547, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8547009229660034, "num_tokens": 692455937.0, "step": 18147 }, { "epoch": 2.3086121358605776, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.789570808410645, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8667942881584167, "num_tokens": 692496277.0, "step": 18148 }, { "epoch": 2.308739346139168, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.778311729431152, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8736555576324463, "num_tokens": 692530485.0, "step": 18149 }, { "epoch": 2.3088665564177586, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.784893035888672, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8695468902587891, "num_tokens": 692564794.0, "step": 18150 }, { "epoch": 2.308993766696349, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.792956352233887, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.862484335899353, "num_tokens": 692610359.0, "step": 18151 }, { "epoch": 2.3091209769749397, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.784192085266113, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8786798715591431, "num_tokens": 692641560.0, "step": 18152 }, { "epoch": 2.30924818725353, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.862504005432129, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8674221038818359, "num_tokens": 692680114.0, "step": 18153 }, { "epoch": 2.3093753975321207, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.76987361907959, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8598593473434448, "num_tokens": 692714199.0, "step": 18154 }, { "epoch": 2.3095026078107113, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.766949653625488, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8748047351837158, "num_tokens": 692752331.0, "step": 18155 }, { "epoch": 2.309629818089302, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.78527545928955, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8784747123718262, "num_tokens": 692789167.0, "step": 18156 }, { "epoch": 2.3097570283678923, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.762434005737305, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8720798492431641, "num_tokens": 692829234.0, "step": 18157 }, { "epoch": 2.3098842386464824, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.802135467529297, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8516994714736938, "num_tokens": 692863749.0, "step": 18158 }, { "epoch": 2.3100114489250734, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.794441223144531, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8676067590713501, "num_tokens": 692900972.0, "step": 18159 }, { "epoch": 2.3101386592036635, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.759735107421875, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8666040301322937, "num_tokens": 692938522.0, "step": 18160 }, { "epoch": 2.310265869482254, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.789961814880371, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.881391704082489, "num_tokens": 692973653.0, "step": 18161 }, { "epoch": 2.3103930797608445, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.795762062072754, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8659461736679077, "num_tokens": 693016325.0, "step": 18162 }, { "epoch": 2.310520290039435, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.745584487915039, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8821684122085571, "num_tokens": 693052022.0, "step": 18163 }, { "epoch": 2.3106475003180256, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.823639869689941, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8770878314971924, "num_tokens": 693093936.0, "step": 18164 }, { "epoch": 2.310774710596616, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.710384368896484, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8853709697723389, "num_tokens": 693130160.0, "step": 18165 }, { "epoch": 2.3109019208752066, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.750580787658691, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.874682605266571, "num_tokens": 693172405.0, "step": 18166 }, { "epoch": 2.311029131153797, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.767648696899414, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8750838041305542, "num_tokens": 693206499.0, "step": 18167 }, { "epoch": 2.3111563414323877, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.775100708007812, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.870697021484375, "num_tokens": 693244287.0, "step": 18168 }, { "epoch": 2.311283551710978, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.761910438537598, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8635339736938477, "num_tokens": 693286164.0, "step": 18169 }, { "epoch": 2.3114107619895687, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.762762069702148, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8785454034805298, "num_tokens": 693330961.0, "step": 18170 }, { "epoch": 2.3115379722681593, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.792641639709473, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8780207633972168, "num_tokens": 693375169.0, "step": 18171 }, { "epoch": 2.31166518254675, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.77786922454834, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8668733835220337, "num_tokens": 693414127.0, "step": 18172 }, { "epoch": 2.3117923928253403, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.816502571105957, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.869013249874115, "num_tokens": 693450446.0, "step": 18173 }, { "epoch": 2.311919603103931, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.808866500854492, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8634976148605347, "num_tokens": 693490973.0, "step": 18174 }, { "epoch": 2.3120468133825214, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.727410316467285, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8741106986999512, "num_tokens": 693526815.0, "step": 18175 }, { "epoch": 2.312174023661112, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.830242156982422, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.865330696105957, "num_tokens": 693566949.0, "step": 18176 }, { "epoch": 2.3123012339397024, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.778273582458496, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8658366203308105, "num_tokens": 693601116.0, "step": 18177 }, { "epoch": 2.312428444218293, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.787559509277344, "learning_rate": 1e-06, "loss": 0.3252, "mean_token_accuracy": 0.9016925096511841, "num_tokens": 693640588.0, "step": 18178 }, { "epoch": 2.3125556544968835, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.807978630065918, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8734232187271118, "num_tokens": 693683352.0, "step": 18179 }, { "epoch": 2.312682864775474, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.799789428710938, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8778522610664368, "num_tokens": 693722579.0, "step": 18180 }, { "epoch": 2.3128100750540646, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.817086219787598, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8730526566505432, "num_tokens": 693766743.0, "step": 18181 }, { "epoch": 2.312937285332655, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.752937316894531, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8749595880508423, "num_tokens": 693809699.0, "step": 18182 }, { "epoch": 2.313064495611245, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.867977142333984, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8781313896179199, "num_tokens": 693845508.0, "step": 18183 }, { "epoch": 2.313191705889836, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.768912315368652, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8718087077140808, "num_tokens": 693883763.0, "step": 18184 }, { "epoch": 2.313318916168426, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.815186500549316, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8777393102645874, "num_tokens": 693917093.0, "step": 18185 }, { "epoch": 2.3134461264470167, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.769787788391113, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8732655644416809, "num_tokens": 693949219.0, "step": 18186 }, { "epoch": 2.3135733367256073, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.721631050109863, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8660258054733276, "num_tokens": 693992755.0, "step": 18187 }, { "epoch": 2.313700547004198, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.820876121520996, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8577446937561035, "num_tokens": 694030386.0, "step": 18188 }, { "epoch": 2.3138277572827883, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.754591941833496, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8803958296775818, "num_tokens": 694065019.0, "step": 18189 }, { "epoch": 2.313954967561379, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.734625816345215, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8702623844146729, "num_tokens": 694106645.0, "step": 18190 }, { "epoch": 2.3140821778399694, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.77093505859375, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8770092725753784, "num_tokens": 694149146.0, "step": 18191 }, { "epoch": 2.31420938811856, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.77554702758789, "learning_rate": 1e-06, "loss": 0.46, "mean_token_accuracy": 0.8560120463371277, "num_tokens": 694193567.0, "step": 18192 }, { "epoch": 2.3143365983971504, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.769521713256836, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8825485706329346, "num_tokens": 694232160.0, "step": 18193 }, { "epoch": 2.314463808675741, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.757229804992676, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8729850053787231, "num_tokens": 694275153.0, "step": 18194 }, { "epoch": 2.3145910189543315, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.749945640563965, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8785158395767212, "num_tokens": 694313722.0, "step": 18195 }, { "epoch": 2.314718229232922, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.773632049560547, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8763450384140015, "num_tokens": 694350544.0, "step": 18196 }, { "epoch": 2.3148454395115126, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.770833015441895, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8738665580749512, "num_tokens": 694386552.0, "step": 18197 }, { "epoch": 2.314972649790103, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.712043762207031, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8690448999404907, "num_tokens": 694434368.0, "step": 18198 }, { "epoch": 2.3150998600686936, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.786587715148926, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8717148303985596, "num_tokens": 694479856.0, "step": 18199 }, { "epoch": 2.315227070347284, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.769247055053711, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8725206851959229, "num_tokens": 694521107.0, "step": 18200 }, { "epoch": 2.3153542806258747, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.817216873168945, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8738358020782471, "num_tokens": 694560886.0, "step": 18201 }, { "epoch": 2.315481490904465, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.786538124084473, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.876830518245697, "num_tokens": 694598404.0, "step": 18202 }, { "epoch": 2.3156087011830557, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.800954818725586, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8679735660552979, "num_tokens": 694637162.0, "step": 18203 }, { "epoch": 2.3157359114616463, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.789621353149414, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8753038644790649, "num_tokens": 694677435.0, "step": 18204 }, { "epoch": 2.315863121740237, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.800914764404297, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8815619945526123, "num_tokens": 694713980.0, "step": 18205 }, { "epoch": 2.3159903320188273, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.873480796813965, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8685235977172852, "num_tokens": 694753884.0, "step": 18206 }, { "epoch": 2.316117542297418, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.766798973083496, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.885369062423706, "num_tokens": 694785739.0, "step": 18207 }, { "epoch": 2.316244752576008, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.838847160339355, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8712040185928345, "num_tokens": 694827629.0, "step": 18208 }, { "epoch": 2.316371962854599, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.83303451538086, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8683083057403564, "num_tokens": 694868241.0, "step": 18209 }, { "epoch": 2.316499173133189, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.8021879196167, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8634421825408936, "num_tokens": 694905298.0, "step": 18210 }, { "epoch": 2.3166263834117795, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.840670585632324, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8651440143585205, "num_tokens": 694938054.0, "step": 18211 }, { "epoch": 2.31675359369037, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.752700805664062, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8685970306396484, "num_tokens": 694978761.0, "step": 18212 }, { "epoch": 2.3168808039689606, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.803288459777832, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8690363168716431, "num_tokens": 695015073.0, "step": 18213 }, { "epoch": 2.317008014247551, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.783595085144043, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8667577505111694, "num_tokens": 695052288.0, "step": 18214 }, { "epoch": 2.3171352245261416, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.739227294921875, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.861506998538971, "num_tokens": 695088976.0, "step": 18215 }, { "epoch": 2.317262434804732, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.809844970703125, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.875605583190918, "num_tokens": 695128482.0, "step": 18216 }, { "epoch": 2.3173896450833227, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.79761791229248, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8689900636672974, "num_tokens": 695164963.0, "step": 18217 }, { "epoch": 2.317516855361913, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.827277183532715, "learning_rate": 1e-06, "loss": 0.3429, "mean_token_accuracy": 0.8940848112106323, "num_tokens": 695201515.0, "step": 18218 }, { "epoch": 2.3176440656405037, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.775208473205566, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8783888816833496, "num_tokens": 695242057.0, "step": 18219 }, { "epoch": 2.3177712759190943, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.851588249206543, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8740925788879395, "num_tokens": 695277128.0, "step": 18220 }, { "epoch": 2.317898486197685, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.821959495544434, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8857650756835938, "num_tokens": 695311984.0, "step": 18221 }, { "epoch": 2.3180256964762753, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.777234077453613, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8626004457473755, "num_tokens": 695348457.0, "step": 18222 }, { "epoch": 2.318152906754866, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.823906898498535, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8748462200164795, "num_tokens": 695386518.0, "step": 18223 }, { "epoch": 2.3182801170334564, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.73824691772461, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8694136142730713, "num_tokens": 695425365.0, "step": 18224 }, { "epoch": 2.318407327312047, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.810603141784668, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.874659538269043, "num_tokens": 695459559.0, "step": 18225 }, { "epoch": 2.3185345375906374, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.757683753967285, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8688076734542847, "num_tokens": 695491386.0, "step": 18226 }, { "epoch": 2.318661747869228, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.741083145141602, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.8556784987449646, "num_tokens": 695534660.0, "step": 18227 }, { "epoch": 2.3187889581478185, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.723297119140625, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8646833896636963, "num_tokens": 695578127.0, "step": 18228 }, { "epoch": 2.318916168426409, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.742753982543945, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8607715964317322, "num_tokens": 695614181.0, "step": 18229 }, { "epoch": 2.3190433787049995, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.814912796020508, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8713997602462769, "num_tokens": 695652753.0, "step": 18230 }, { "epoch": 2.3191705889835896, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.775810241699219, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.879202127456665, "num_tokens": 695692449.0, "step": 18231 }, { "epoch": 2.3192977992621806, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.760855674743652, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8718984127044678, "num_tokens": 695730812.0, "step": 18232 }, { "epoch": 2.3194250095407707, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.861625671386719, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8873917460441589, "num_tokens": 695767577.0, "step": 18233 }, { "epoch": 2.319552219819361, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.771428108215332, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8726090788841248, "num_tokens": 695812372.0, "step": 18234 }, { "epoch": 2.3196794300979517, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.830766677856445, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8867353200912476, "num_tokens": 695847470.0, "step": 18235 }, { "epoch": 2.3198066403765423, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.812320709228516, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8882660269737244, "num_tokens": 695888082.0, "step": 18236 }, { "epoch": 2.319933850655133, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.792135238647461, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8806377649307251, "num_tokens": 695928904.0, "step": 18237 }, { "epoch": 2.3200610609337233, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.818284034729004, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8617952466011047, "num_tokens": 695967549.0, "step": 18238 }, { "epoch": 2.320188271212314, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.768299102783203, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8665218353271484, "num_tokens": 696005611.0, "step": 18239 }, { "epoch": 2.3203154814909044, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.812337875366211, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8755624294281006, "num_tokens": 696034864.0, "step": 18240 }, { "epoch": 2.320442691769495, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.795462608337402, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8707411289215088, "num_tokens": 696072137.0, "step": 18241 }, { "epoch": 2.3205699020480854, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.792946815490723, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8746413588523865, "num_tokens": 696115744.0, "step": 18242 }, { "epoch": 2.320697112326676, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.801789283752441, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8685693740844727, "num_tokens": 696151293.0, "step": 18243 }, { "epoch": 2.3208243226052665, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.810795783996582, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.8898274898529053, "num_tokens": 696190742.0, "step": 18244 }, { "epoch": 2.320951532883857, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.780884742736816, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.861759603023529, "num_tokens": 696232807.0, "step": 18245 }, { "epoch": 2.3210787431624476, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.795419692993164, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8691789507865906, "num_tokens": 696269416.0, "step": 18246 }, { "epoch": 2.321205953441038, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.778767585754395, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8737348318099976, "num_tokens": 696305144.0, "step": 18247 }, { "epoch": 2.3213331637196286, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.787100791931152, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8645082712173462, "num_tokens": 696344854.0, "step": 18248 }, { "epoch": 2.321460373998219, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.833337783813477, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8907943964004517, "num_tokens": 696380154.0, "step": 18249 }, { "epoch": 2.3215875842768097, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.811131477355957, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8591639995574951, "num_tokens": 696416458.0, "step": 18250 }, { "epoch": 2.3217147945554, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.787063598632812, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8753353357315063, "num_tokens": 696453232.0, "step": 18251 }, { "epoch": 2.3218420048339907, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.852725982666016, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8529162406921387, "num_tokens": 696487930.0, "step": 18252 }, { "epoch": 2.3219692151125813, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.814703941345215, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8637077808380127, "num_tokens": 696523824.0, "step": 18253 }, { "epoch": 2.322096425391172, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.792271614074707, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8818598985671997, "num_tokens": 696567007.0, "step": 18254 }, { "epoch": 2.3222236356697623, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.793601989746094, "learning_rate": 1e-06, "loss": 0.3582, "mean_token_accuracy": 0.8879413604736328, "num_tokens": 696610055.0, "step": 18255 }, { "epoch": 2.3223508459483524, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.853824615478516, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8581610918045044, "num_tokens": 696644058.0, "step": 18256 }, { "epoch": 2.3224780562269434, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.777193069458008, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8804003000259399, "num_tokens": 696686381.0, "step": 18257 }, { "epoch": 2.3226052665055335, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.837340354919434, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8666871786117554, "num_tokens": 696727567.0, "step": 18258 }, { "epoch": 2.322732476784124, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.793092727661133, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8624805808067322, "num_tokens": 696767266.0, "step": 18259 }, { "epoch": 2.3228596870627145, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.789756774902344, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8856804370880127, "num_tokens": 696806059.0, "step": 18260 }, { "epoch": 2.322986897341305, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.806736946105957, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8711056709289551, "num_tokens": 696844232.0, "step": 18261 }, { "epoch": 2.3231141076198956, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.798409461975098, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8751671314239502, "num_tokens": 696886223.0, "step": 18262 }, { "epoch": 2.323241317898486, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.773083686828613, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8591077923774719, "num_tokens": 696922625.0, "step": 18263 }, { "epoch": 2.3233685281770766, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.817906379699707, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8615199327468872, "num_tokens": 696965160.0, "step": 18264 }, { "epoch": 2.323495738455667, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.764521598815918, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8772958517074585, "num_tokens": 697000854.0, "step": 18265 }, { "epoch": 2.3236229487342577, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.756850242614746, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.8864732980728149, "num_tokens": 697041264.0, "step": 18266 }, { "epoch": 2.323750159012848, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.7959566116333, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8697463274002075, "num_tokens": 697080079.0, "step": 18267 }, { "epoch": 2.3238773692914387, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.786656379699707, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8694062232971191, "num_tokens": 697123755.0, "step": 18268 }, { "epoch": 2.3240045795700293, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.830740928649902, "learning_rate": 1e-06, "loss": 0.4949, "mean_token_accuracy": 0.8444185853004456, "num_tokens": 697162012.0, "step": 18269 }, { "epoch": 2.32413178984862, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.832436561584473, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8815996646881104, "num_tokens": 697200188.0, "step": 18270 }, { "epoch": 2.3242590001272103, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.76708698272705, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8747016787528992, "num_tokens": 697240536.0, "step": 18271 }, { "epoch": 2.324386210405801, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.796770095825195, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8837621808052063, "num_tokens": 697279459.0, "step": 18272 }, { "epoch": 2.3245134206843914, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.797745704650879, "learning_rate": 1e-06, "loss": 0.3439, "mean_token_accuracy": 0.8894047141075134, "num_tokens": 697312594.0, "step": 18273 }, { "epoch": 2.324640630962982, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.79394817352295, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8598737716674805, "num_tokens": 697357834.0, "step": 18274 }, { "epoch": 2.3247678412415724, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.860711097717285, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8712602853775024, "num_tokens": 697401522.0, "step": 18275 }, { "epoch": 2.324895051520163, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.781560897827148, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8701401948928833, "num_tokens": 697440415.0, "step": 18276 }, { "epoch": 2.3250222617987535, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.837638854980469, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8566079139709473, "num_tokens": 697477609.0, "step": 18277 }, { "epoch": 2.325149472077344, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.744791984558105, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.868567168712616, "num_tokens": 697515566.0, "step": 18278 }, { "epoch": 2.3252766823559345, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.788013458251953, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8760813474655151, "num_tokens": 697549537.0, "step": 18279 }, { "epoch": 2.325403892634525, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.786519050598145, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8770408034324646, "num_tokens": 697582618.0, "step": 18280 }, { "epoch": 2.325531102913115, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.777681350708008, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8851575255393982, "num_tokens": 697622234.0, "step": 18281 }, { "epoch": 2.325658313191706, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.811121940612793, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8646801710128784, "num_tokens": 697657504.0, "step": 18282 }, { "epoch": 2.325785523470296, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.839210510253906, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.859657883644104, "num_tokens": 697699496.0, "step": 18283 }, { "epoch": 2.3259127337488867, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.718501091003418, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8848084807395935, "num_tokens": 697739969.0, "step": 18284 }, { "epoch": 2.3260399440274773, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.850652694702148, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8627256751060486, "num_tokens": 697774346.0, "step": 18285 }, { "epoch": 2.326167154306068, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.729290008544922, "learning_rate": 1e-06, "loss": 0.4475, "mean_token_accuracy": 0.8580443263053894, "num_tokens": 697813013.0, "step": 18286 }, { "epoch": 2.3262943645846583, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.837751388549805, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8659056425094604, "num_tokens": 697854163.0, "step": 18287 }, { "epoch": 2.326421574863249, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.776227951049805, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8798332214355469, "num_tokens": 697898049.0, "step": 18288 }, { "epoch": 2.3265487851418394, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.839201927185059, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8753474950790405, "num_tokens": 697936086.0, "step": 18289 }, { "epoch": 2.32667599542043, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.763161659240723, "learning_rate": 1e-06, "loss": 0.4957, "mean_token_accuracy": 0.8486382961273193, "num_tokens": 697981529.0, "step": 18290 }, { "epoch": 2.3268032056990204, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.86618423461914, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8895078897476196, "num_tokens": 698016047.0, "step": 18291 }, { "epoch": 2.326930415977611, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.726212501525879, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8844588398933411, "num_tokens": 698052408.0, "step": 18292 }, { "epoch": 2.3270576262562015, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.858160972595215, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8654208183288574, "num_tokens": 698089751.0, "step": 18293 }, { "epoch": 2.327184836534792, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.766104698181152, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8711813688278198, "num_tokens": 698121274.0, "step": 18294 }, { "epoch": 2.3273120468133826, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.799031257629395, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8808520436286926, "num_tokens": 698155880.0, "step": 18295 }, { "epoch": 2.327439257091973, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.843327522277832, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8745670318603516, "num_tokens": 698191448.0, "step": 18296 }, { "epoch": 2.3275664673705636, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.822439193725586, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8675907850265503, "num_tokens": 698228437.0, "step": 18297 }, { "epoch": 2.327693677649154, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.833680152893066, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8649926781654358, "num_tokens": 698268714.0, "step": 18298 }, { "epoch": 2.3278208879277447, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.840577125549316, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.872770369052887, "num_tokens": 698305575.0, "step": 18299 }, { "epoch": 2.327948098206335, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.874591827392578, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8677560091018677, "num_tokens": 698347701.0, "step": 18300 }, { "epoch": 2.3280753084849257, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.808425903320312, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8875473141670227, "num_tokens": 698385906.0, "step": 18301 }, { "epoch": 2.3282025187635162, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.861493110656738, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.871741533279419, "num_tokens": 698415511.0, "step": 18302 }, { "epoch": 2.3283297290421068, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.864871978759766, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8677030801773071, "num_tokens": 698460528.0, "step": 18303 }, { "epoch": 2.3284569393206973, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.869240760803223, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.858222246170044, "num_tokens": 698500800.0, "step": 18304 }, { "epoch": 2.328584149599288, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.816527366638184, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8711032867431641, "num_tokens": 698539373.0, "step": 18305 }, { "epoch": 2.328711359877878, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.835020065307617, "learning_rate": 1e-06, "loss": 0.3477, "mean_token_accuracy": 0.8944033980369568, "num_tokens": 698577067.0, "step": 18306 }, { "epoch": 2.328838570156469, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.800101280212402, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8716037273406982, "num_tokens": 698613515.0, "step": 18307 }, { "epoch": 2.328965780435059, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.81428050994873, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8741275668144226, "num_tokens": 698647278.0, "step": 18308 }, { "epoch": 2.3290929907136495, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.810128211975098, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8800827264785767, "num_tokens": 698688738.0, "step": 18309 }, { "epoch": 2.32922020099224, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.794894218444824, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8761500120162964, "num_tokens": 698725003.0, "step": 18310 }, { "epoch": 2.3293474112708306, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.793643951416016, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8544648885726929, "num_tokens": 698768205.0, "step": 18311 }, { "epoch": 2.329474621549421, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.804830551147461, "learning_rate": 1e-06, "loss": 0.4636, "mean_token_accuracy": 0.8535560965538025, "num_tokens": 698805352.0, "step": 18312 }, { "epoch": 2.3296018318280116, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.753339767456055, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.881938099861145, "num_tokens": 698836610.0, "step": 18313 }, { "epoch": 2.329729042106602, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.869735717773438, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8818634152412415, "num_tokens": 698877514.0, "step": 18314 }, { "epoch": 2.3298562523851927, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.799880981445312, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8754276037216187, "num_tokens": 698916699.0, "step": 18315 }, { "epoch": 2.329983462663783, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.809103012084961, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8725345134735107, "num_tokens": 698953568.0, "step": 18316 }, { "epoch": 2.3301106729423737, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.861783027648926, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8670080900192261, "num_tokens": 698990562.0, "step": 18317 }, { "epoch": 2.3302378832209643, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.835315704345703, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8748703002929688, "num_tokens": 699024095.0, "step": 18318 }, { "epoch": 2.330365093499555, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.800684928894043, "learning_rate": 1e-06, "loss": 0.4598, "mean_token_accuracy": 0.8531102538108826, "num_tokens": 699062327.0, "step": 18319 }, { "epoch": 2.3304923037781453, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.840174674987793, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8769767880439758, "num_tokens": 699101591.0, "step": 18320 }, { "epoch": 2.330619514056736, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.8298978805542, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8644536733627319, "num_tokens": 699134836.0, "step": 18321 }, { "epoch": 2.3307467243353264, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.81544017791748, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8631187677383423, "num_tokens": 699171218.0, "step": 18322 }, { "epoch": 2.330873934613917, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.827682495117188, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8898404836654663, "num_tokens": 699200891.0, "step": 18323 }, { "epoch": 2.3310011448925074, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.864461898803711, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8877017498016357, "num_tokens": 699238080.0, "step": 18324 }, { "epoch": 2.331128355171098, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.847400665283203, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8609796166419983, "num_tokens": 699276051.0, "step": 18325 }, { "epoch": 2.3312555654496885, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.761777877807617, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8792526125907898, "num_tokens": 699315721.0, "step": 18326 }, { "epoch": 2.331382775728279, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.78604793548584, "learning_rate": 1e-06, "loss": 0.3579, "mean_token_accuracy": 0.8846172094345093, "num_tokens": 699350342.0, "step": 18327 }, { "epoch": 2.3315099860068695, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.824335098266602, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8785660862922668, "num_tokens": 699390301.0, "step": 18328 }, { "epoch": 2.3316371962854596, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.857627868652344, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8582173585891724, "num_tokens": 699427135.0, "step": 18329 }, { "epoch": 2.3317644065640506, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.763792991638184, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8777759075164795, "num_tokens": 699466461.0, "step": 18330 }, { "epoch": 2.3318916168426407, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.76141357421875, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8655612468719482, "num_tokens": 699502446.0, "step": 18331 }, { "epoch": 2.332018827121231, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.800889015197754, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.872545063495636, "num_tokens": 699546357.0, "step": 18332 }, { "epoch": 2.3321460373998217, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.767229080200195, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8718217015266418, "num_tokens": 699581879.0, "step": 18333 }, { "epoch": 2.3322732476784123, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.872367858886719, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8861815333366394, "num_tokens": 699616287.0, "step": 18334 }, { "epoch": 2.332400457957003, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.75256633758545, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8708701133728027, "num_tokens": 699655409.0, "step": 18335 }, { "epoch": 2.3325276682355933, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.816003799438477, "learning_rate": 1e-06, "loss": 0.4759, "mean_token_accuracy": 0.846875011920929, "num_tokens": 699701420.0, "step": 18336 }, { "epoch": 2.332654878514184, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.861222267150879, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8744761347770691, "num_tokens": 699736151.0, "step": 18337 }, { "epoch": 2.3327820887927744, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.794326782226562, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8657341003417969, "num_tokens": 699770111.0, "step": 18338 }, { "epoch": 2.332909299071365, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.831563949584961, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8686105012893677, "num_tokens": 699807442.0, "step": 18339 }, { "epoch": 2.3330365093499554, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.79723072052002, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.870299220085144, "num_tokens": 699849364.0, "step": 18340 }, { "epoch": 2.333163719628546, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.790434837341309, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8869826793670654, "num_tokens": 699886234.0, "step": 18341 }, { "epoch": 2.3332909299071365, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.785012245178223, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.8551074862480164, "num_tokens": 699925660.0, "step": 18342 }, { "epoch": 2.333418140185727, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.862255096435547, "learning_rate": 1e-06, "loss": 0.4868, "mean_token_accuracy": 0.8456971645355225, "num_tokens": 699967894.0, "step": 18343 }, { "epoch": 2.3335453504643175, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.805419921875, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8670607805252075, "num_tokens": 699999211.0, "step": 18344 }, { "epoch": 2.333672560742908, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.791666984558105, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8799446821212769, "num_tokens": 700037639.0, "step": 18345 }, { "epoch": 2.3337997710214986, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.784924507141113, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8652390837669373, "num_tokens": 700075677.0, "step": 18346 }, { "epoch": 2.333926981300089, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.883127212524414, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8787904977798462, "num_tokens": 700115697.0, "step": 18347 }, { "epoch": 2.3340541915786797, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.765302658081055, "learning_rate": 1e-06, "loss": 0.4489, "mean_token_accuracy": 0.8547442555427551, "num_tokens": 700152553.0, "step": 18348 }, { "epoch": 2.33418140185727, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.792877197265625, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8784021735191345, "num_tokens": 700190958.0, "step": 18349 }, { "epoch": 2.3343086121358607, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.798925399780273, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8702021837234497, "num_tokens": 700224244.0, "step": 18350 }, { "epoch": 2.3344358224144512, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.834239959716797, "learning_rate": 1e-06, "loss": 0.3242, "mean_token_accuracy": 0.8977806568145752, "num_tokens": 700259736.0, "step": 18351 }, { "epoch": 2.3345630326930418, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.842659950256348, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.872268795967102, "num_tokens": 700293624.0, "step": 18352 }, { "epoch": 2.3346902429716323, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.747032165527344, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8760616183280945, "num_tokens": 700328052.0, "step": 18353 }, { "epoch": 2.3348174532502224, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.855592727661133, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8651858568191528, "num_tokens": 700364652.0, "step": 18354 }, { "epoch": 2.3349446635288134, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.819018363952637, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8848433494567871, "num_tokens": 700403698.0, "step": 18355 }, { "epoch": 2.3350718738074034, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.840688705444336, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8859479427337646, "num_tokens": 700439391.0, "step": 18356 }, { "epoch": 2.335199084085994, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.829222679138184, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8750830888748169, "num_tokens": 700475095.0, "step": 18357 }, { "epoch": 2.3353262943645845, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.843273162841797, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8765541315078735, "num_tokens": 700514167.0, "step": 18358 }, { "epoch": 2.335453504643175, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.833041191101074, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8807632327079773, "num_tokens": 700548333.0, "step": 18359 }, { "epoch": 2.3355807149217656, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.842118263244629, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8773810863494873, "num_tokens": 700594261.0, "step": 18360 }, { "epoch": 2.335707925200356, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.90494155883789, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.877996563911438, "num_tokens": 700632513.0, "step": 18361 }, { "epoch": 2.3358351354789466, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.796656608581543, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8788143992424011, "num_tokens": 700672430.0, "step": 18362 }, { "epoch": 2.335962345757537, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.853816032409668, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8782769441604614, "num_tokens": 700714321.0, "step": 18363 }, { "epoch": 2.3360895560361277, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.892989158630371, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8726787567138672, "num_tokens": 700751714.0, "step": 18364 }, { "epoch": 2.336216766314718, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.861698150634766, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8698514699935913, "num_tokens": 700794087.0, "step": 18365 }, { "epoch": 2.3363439765933087, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.83250904083252, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8681530952453613, "num_tokens": 700832031.0, "step": 18366 }, { "epoch": 2.3364711868718993, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.814682006835938, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8752388954162598, "num_tokens": 700867175.0, "step": 18367 }, { "epoch": 2.33659839715049, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.825997352600098, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8786290884017944, "num_tokens": 700906762.0, "step": 18368 }, { "epoch": 2.3367256074290803, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.832693099975586, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8754185438156128, "num_tokens": 700944316.0, "step": 18369 }, { "epoch": 2.336852817707671, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.801597595214844, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8680521249771118, "num_tokens": 700979369.0, "step": 18370 }, { "epoch": 2.3369800279862614, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.899762153625488, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8691127300262451, "num_tokens": 701016213.0, "step": 18371 }, { "epoch": 2.337107238264852, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.771503448486328, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8708177208900452, "num_tokens": 701055270.0, "step": 18372 }, { "epoch": 2.3372344485434424, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.855822563171387, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8713454604148865, "num_tokens": 701091058.0, "step": 18373 }, { "epoch": 2.337361658822033, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.809029579162598, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8806923627853394, "num_tokens": 701129446.0, "step": 18374 }, { "epoch": 2.3374888691006235, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.834149360656738, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8757693767547607, "num_tokens": 701168726.0, "step": 18375 }, { "epoch": 2.337616079379214, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.770090103149414, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8672285079956055, "num_tokens": 701204041.0, "step": 18376 }, { "epoch": 2.3377432896578045, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.876089096069336, "learning_rate": 1e-06, "loss": 0.4685, "mean_token_accuracy": 0.8502137660980225, "num_tokens": 701244660.0, "step": 18377 }, { "epoch": 2.337870499936395, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.828573226928711, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.864590048789978, "num_tokens": 701284830.0, "step": 18378 }, { "epoch": 2.337997710214985, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.827131271362305, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8796112537384033, "num_tokens": 701319720.0, "step": 18379 }, { "epoch": 2.338124920493576, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.803317070007324, "learning_rate": 1e-06, "loss": 0.4691, "mean_token_accuracy": 0.8516581058502197, "num_tokens": 701361770.0, "step": 18380 }, { "epoch": 2.338252130772166, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.8286714553833, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8694256544113159, "num_tokens": 701399777.0, "step": 18381 }, { "epoch": 2.3383793410507567, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.837271690368652, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8722387552261353, "num_tokens": 701432680.0, "step": 18382 }, { "epoch": 2.3385065513293473, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.843629837036133, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8830867409706116, "num_tokens": 701474101.0, "step": 18383 }, { "epoch": 2.338633761607938, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.826659202575684, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8857772946357727, "num_tokens": 701517568.0, "step": 18384 }, { "epoch": 2.3387609718865283, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.820266723632812, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8655389547348022, "num_tokens": 701558466.0, "step": 18385 }, { "epoch": 2.338888182165119, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.898714065551758, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8676677346229553, "num_tokens": 701591104.0, "step": 18386 }, { "epoch": 2.3390153924437094, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.795169830322266, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8657814860343933, "num_tokens": 701633242.0, "step": 18387 }, { "epoch": 2.3391426027223, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.847136497497559, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8714481592178345, "num_tokens": 701668864.0, "step": 18388 }, { "epoch": 2.3392698130008904, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.852710723876953, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8768459558486938, "num_tokens": 701703648.0, "step": 18389 }, { "epoch": 2.339397023279481, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.865691184997559, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8685596585273743, "num_tokens": 701744417.0, "step": 18390 }, { "epoch": 2.3395242335580715, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.818161010742188, "learning_rate": 1e-06, "loss": 0.4876, "mean_token_accuracy": 0.8406295776367188, "num_tokens": 701783541.0, "step": 18391 }, { "epoch": 2.339651443836662, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.885786056518555, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8687700033187866, "num_tokens": 701825054.0, "step": 18392 }, { "epoch": 2.3397786541152525, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.827055931091309, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8793526887893677, "num_tokens": 701866173.0, "step": 18393 }, { "epoch": 2.339905864393843, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.837803840637207, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.868260383605957, "num_tokens": 701904812.0, "step": 18394 }, { "epoch": 2.3400330746724336, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.831863403320312, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.862565815448761, "num_tokens": 701942590.0, "step": 18395 }, { "epoch": 2.340160284951024, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.855206489562988, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.875584065914154, "num_tokens": 701980773.0, "step": 18396 }, { "epoch": 2.3402874952296147, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.897168159484863, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8854040503501892, "num_tokens": 702014508.0, "step": 18397 }, { "epoch": 2.340414705508205, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.853261947631836, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8847860097885132, "num_tokens": 702055754.0, "step": 18398 }, { "epoch": 2.3405419157867957, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.805959701538086, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8708181381225586, "num_tokens": 702100022.0, "step": 18399 }, { "epoch": 2.3406691260653862, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.903885841369629, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8619797229766846, "num_tokens": 702133217.0, "step": 18400 }, { "epoch": 2.3407963363439768, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.766495704650879, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8851820230484009, "num_tokens": 702176998.0, "step": 18401 }, { "epoch": 2.3409235466225673, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.898204803466797, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8790981769561768, "num_tokens": 702216219.0, "step": 18402 }, { "epoch": 2.341050756901158, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.858983039855957, "learning_rate": 1e-06, "loss": 0.4452, "mean_token_accuracy": 0.8673439621925354, "num_tokens": 702246502.0, "step": 18403 }, { "epoch": 2.341177967179748, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.860123634338379, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8796880841255188, "num_tokens": 702277990.0, "step": 18404 }, { "epoch": 2.341305177458339, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.850640296936035, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8751788139343262, "num_tokens": 702312742.0, "step": 18405 }, { "epoch": 2.341432387736929, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.797286033630371, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8871089816093445, "num_tokens": 702348575.0, "step": 18406 }, { "epoch": 2.3415595980155195, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.879714965820312, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8613282442092896, "num_tokens": 702385956.0, "step": 18407 }, { "epoch": 2.34168680829411, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.793599128723145, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8774696588516235, "num_tokens": 702423973.0, "step": 18408 }, { "epoch": 2.3418140185727006, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.900513648986816, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8698756098747253, "num_tokens": 702455508.0, "step": 18409 }, { "epoch": 2.341941228851291, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.844112396240234, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8826924562454224, "num_tokens": 702493365.0, "step": 18410 }, { "epoch": 2.3420684391298816, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.86600399017334, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8742997050285339, "num_tokens": 702528036.0, "step": 18411 }, { "epoch": 2.342195649408472, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.871613502502441, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8675752282142639, "num_tokens": 702571238.0, "step": 18412 }, { "epoch": 2.3423228596870627, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.825934410095215, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8685286045074463, "num_tokens": 702611247.0, "step": 18413 }, { "epoch": 2.342450069965653, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.858831405639648, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8732578754425049, "num_tokens": 702649661.0, "step": 18414 }, { "epoch": 2.3425772802442437, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.824257850646973, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8659050464630127, "num_tokens": 702693573.0, "step": 18415 }, { "epoch": 2.3427044905228342, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.870644569396973, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8836108446121216, "num_tokens": 702735351.0, "step": 18416 }, { "epoch": 2.3428317008014248, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.819158554077148, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8783565759658813, "num_tokens": 702773406.0, "step": 18417 }, { "epoch": 2.3429589110800153, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.868721961975098, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.879129946231842, "num_tokens": 702809939.0, "step": 18418 }, { "epoch": 2.343086121358606, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.833280563354492, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8703774809837341, "num_tokens": 702851120.0, "step": 18419 }, { "epoch": 2.3432133316371964, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.84544849395752, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8814219832420349, "num_tokens": 702889492.0, "step": 18420 }, { "epoch": 2.343340541915787, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.858903884887695, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8668932914733887, "num_tokens": 702930486.0, "step": 18421 }, { "epoch": 2.3434677521943774, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.783161163330078, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8647981882095337, "num_tokens": 702969866.0, "step": 18422 }, { "epoch": 2.343594962472968, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.827917098999023, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8645268678665161, "num_tokens": 703013703.0, "step": 18423 }, { "epoch": 2.3437221727515585, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.826937675476074, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8620602488517761, "num_tokens": 703054849.0, "step": 18424 }, { "epoch": 2.343849383030149, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.866507530212402, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8678172826766968, "num_tokens": 703088250.0, "step": 18425 }, { "epoch": 2.3439765933087395, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.834068298339844, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8691209554672241, "num_tokens": 703124072.0, "step": 18426 }, { "epoch": 2.3441038035873296, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.857402801513672, "learning_rate": 1e-06, "loss": 0.4562, "mean_token_accuracy": 0.8547896146774292, "num_tokens": 703158869.0, "step": 18427 }, { "epoch": 2.3442310138659206, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.867883682250977, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8745107650756836, "num_tokens": 703193342.0, "step": 18428 }, { "epoch": 2.3443582241445107, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.88333797454834, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8852070569992065, "num_tokens": 703229899.0, "step": 18429 }, { "epoch": 2.344485434423101, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.805388450622559, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8713732361793518, "num_tokens": 703266735.0, "step": 18430 }, { "epoch": 2.3446126447016917, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.876194953918457, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8689494132995605, "num_tokens": 703307515.0, "step": 18431 }, { "epoch": 2.3447398549802823, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.848155975341797, "learning_rate": 1e-06, "loss": 0.3549, "mean_token_accuracy": 0.8831812143325806, "num_tokens": 703341237.0, "step": 18432 }, { "epoch": 2.344867065258873, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.867719650268555, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8827738165855408, "num_tokens": 703378900.0, "step": 18433 }, { "epoch": 2.3449942755374633, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.878116607666016, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8808966279029846, "num_tokens": 703415260.0, "step": 18434 }, { "epoch": 2.345121485816054, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.854474067687988, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8775432109832764, "num_tokens": 703452399.0, "step": 18435 }, { "epoch": 2.3452486960946444, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.837821006774902, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8738746643066406, "num_tokens": 703492700.0, "step": 18436 }, { "epoch": 2.345375906373235, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.800827026367188, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8834443092346191, "num_tokens": 703530805.0, "step": 18437 }, { "epoch": 2.3455031166518254, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.87255859375, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8815690279006958, "num_tokens": 703571894.0, "step": 18438 }, { "epoch": 2.345630326930416, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.821945190429688, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8781521320343018, "num_tokens": 703612583.0, "step": 18439 }, { "epoch": 2.3457575372090065, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.859781265258789, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.866011381149292, "num_tokens": 703649100.0, "step": 18440 }, { "epoch": 2.345884747487597, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.85025405883789, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8873674869537354, "num_tokens": 703680425.0, "step": 18441 }, { "epoch": 2.3460119577661875, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.838330268859863, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8785759210586548, "num_tokens": 703721819.0, "step": 18442 }, { "epoch": 2.346139168044778, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.86644458770752, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8688870668411255, "num_tokens": 703755432.0, "step": 18443 }, { "epoch": 2.3462663783233686, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.805953979492188, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8714694380760193, "num_tokens": 703795903.0, "step": 18444 }, { "epoch": 2.346393588601959, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.88294792175293, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8844282031059265, "num_tokens": 703836074.0, "step": 18445 }, { "epoch": 2.3465207988805497, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.80769157409668, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8774516582489014, "num_tokens": 703870609.0, "step": 18446 }, { "epoch": 2.34664800915914, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.893528938293457, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8784222602844238, "num_tokens": 703909034.0, "step": 18447 }, { "epoch": 2.3467752194377307, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.825441360473633, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.865143895149231, "num_tokens": 703946976.0, "step": 18448 }, { "epoch": 2.3469024297163212, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.793457984924316, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8753926753997803, "num_tokens": 703986379.0, "step": 18449 }, { "epoch": 2.3470296399949118, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.850464820861816, "learning_rate": 1e-06, "loss": 0.342, "mean_token_accuracy": 0.8920736312866211, "num_tokens": 704019461.0, "step": 18450 }, { "epoch": 2.3471568502735023, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.81761646270752, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8598891496658325, "num_tokens": 704053411.0, "step": 18451 }, { "epoch": 2.3472840605520924, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.8311185836792, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8777523636817932, "num_tokens": 704092633.0, "step": 18452 }, { "epoch": 2.3474112708306833, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.823375701904297, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8775246143341064, "num_tokens": 704126955.0, "step": 18453 }, { "epoch": 2.3475384811092734, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.79358196258545, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8898262977600098, "num_tokens": 704162222.0, "step": 18454 }, { "epoch": 2.347665691387864, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.831107139587402, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8739925622940063, "num_tokens": 704198949.0, "step": 18455 }, { "epoch": 2.3477929016664545, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.893542289733887, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.868236243724823, "num_tokens": 704235327.0, "step": 18456 }, { "epoch": 2.347920111945045, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.765084266662598, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8796705603599548, "num_tokens": 704272783.0, "step": 18457 }, { "epoch": 2.3480473222236355, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.816338539123535, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8568159937858582, "num_tokens": 704314396.0, "step": 18458 }, { "epoch": 2.348174532502226, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 13.801385879516602, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.8896422982215881, "num_tokens": 704346903.0, "step": 18459 }, { "epoch": 2.3483017427808166, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.828794479370117, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8810704350471497, "num_tokens": 704380456.0, "step": 18460 }, { "epoch": 2.348428953059407, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.840261459350586, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8852090835571289, "num_tokens": 704416403.0, "step": 18461 }, { "epoch": 2.3485561633379977, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.767088890075684, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8745736479759216, "num_tokens": 704460431.0, "step": 18462 }, { "epoch": 2.348683373616588, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.815604209899902, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8817158341407776, "num_tokens": 704497872.0, "step": 18463 }, { "epoch": 2.3488105838951787, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.830013275146484, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.875635027885437, "num_tokens": 704538180.0, "step": 18464 }, { "epoch": 2.3489377941737692, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.788009643554688, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8850587010383606, "num_tokens": 704578342.0, "step": 18465 }, { "epoch": 2.3490650044523598, "ewc_loss": 0.0361328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 13.812602043151855, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8847898840904236, "num_tokens": 704611792.0, "step": 18466 }, { "epoch": 2.3491922147309503, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.828357696533203, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.875730574131012, "num_tokens": 704649898.0, "step": 18467 }, { "epoch": 2.349319425009541, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.821824073791504, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8819280862808228, "num_tokens": 704690632.0, "step": 18468 }, { "epoch": 2.3494466352881314, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.805849075317383, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8640864491462708, "num_tokens": 704722955.0, "step": 18469 }, { "epoch": 2.349573845566722, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.842656135559082, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8736929297447205, "num_tokens": 704762678.0, "step": 18470 }, { "epoch": 2.3497010558453124, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.83411979675293, "learning_rate": 1e-06, "loss": 0.3239, "mean_token_accuracy": 0.8988367319107056, "num_tokens": 704805440.0, "step": 18471 }, { "epoch": 2.349828266123903, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.851765632629395, "learning_rate": 1e-06, "loss": 0.463, "mean_token_accuracy": 0.8546221852302551, "num_tokens": 704848153.0, "step": 18472 }, { "epoch": 2.3499554764024935, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.814236640930176, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8712108135223389, "num_tokens": 704886739.0, "step": 18473 }, { "epoch": 2.350082686681084, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.833806991577148, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8729546070098877, "num_tokens": 704926145.0, "step": 18474 }, { "epoch": 2.3502098969596745, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.798089981079102, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8586007356643677, "num_tokens": 704964135.0, "step": 18475 }, { "epoch": 2.350337107238265, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.87864875793457, "learning_rate": 1e-06, "loss": 0.4497, "mean_token_accuracy": 0.862442672252655, "num_tokens": 705002171.0, "step": 18476 }, { "epoch": 2.350464317516855, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.828140258789062, "learning_rate": 1e-06, "loss": 0.336, "mean_token_accuracy": 0.8932733535766602, "num_tokens": 705040660.0, "step": 18477 }, { "epoch": 2.350591527795446, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.78190803527832, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8800990581512451, "num_tokens": 705073655.0, "step": 18478 }, { "epoch": 2.350718738074036, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.828422546386719, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8705317378044128, "num_tokens": 705115497.0, "step": 18479 }, { "epoch": 2.3508459483526267, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.83923053741455, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8903687000274658, "num_tokens": 705152642.0, "step": 18480 }, { "epoch": 2.3509731586312173, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.792398452758789, "learning_rate": 1e-06, "loss": 0.4549, "mean_token_accuracy": 0.8596091866493225, "num_tokens": 705187282.0, "step": 18481 }, { "epoch": 2.351100368909808, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.876218795776367, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8771457672119141, "num_tokens": 705225635.0, "step": 18482 }, { "epoch": 2.3512275791883983, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.818777084350586, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8826079368591309, "num_tokens": 705260376.0, "step": 18483 }, { "epoch": 2.351354789466989, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.854229927062988, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8727964758872986, "num_tokens": 705298799.0, "step": 18484 }, { "epoch": 2.3514819997455794, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.903130531311035, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8564584851264954, "num_tokens": 705332570.0, "step": 18485 }, { "epoch": 2.35160921002417, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.849064826965332, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8758420348167419, "num_tokens": 705371214.0, "step": 18486 }, { "epoch": 2.3517364203027604, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.869787216186523, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8707528114318848, "num_tokens": 705405051.0, "step": 18487 }, { "epoch": 2.351863630581351, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.82580280303955, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8720319271087646, "num_tokens": 705442460.0, "step": 18488 }, { "epoch": 2.3519908408599415, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.894927024841309, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8730126619338989, "num_tokens": 705482929.0, "step": 18489 }, { "epoch": 2.352118051138532, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.779120445251465, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8689368963241577, "num_tokens": 705525313.0, "step": 18490 }, { "epoch": 2.3522452614171225, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.864142417907715, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8686002492904663, "num_tokens": 705563730.0, "step": 18491 }, { "epoch": 2.352372471695713, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.812753677368164, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8772934675216675, "num_tokens": 705602459.0, "step": 18492 }, { "epoch": 2.3524996819743036, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.820723533630371, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.870770275592804, "num_tokens": 705636223.0, "step": 18493 }, { "epoch": 2.352626892252894, "ewc_loss": 0.036376953125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.78201961517334, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8725386261940002, "num_tokens": 705676178.0, "step": 18494 }, { "epoch": 2.3527541025314846, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.789857864379883, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8801521062850952, "num_tokens": 705716687.0, "step": 18495 }, { "epoch": 2.352881312810075, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.80894660949707, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.88032466173172, "num_tokens": 705759163.0, "step": 18496 }, { "epoch": 2.3530085230886657, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.813031196594238, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8760232925415039, "num_tokens": 705795982.0, "step": 18497 }, { "epoch": 2.3531357333672562, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.810388565063477, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8780558109283447, "num_tokens": 705835601.0, "step": 18498 }, { "epoch": 2.3532629436458468, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.771881103515625, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8717467188835144, "num_tokens": 705876858.0, "step": 18499 }, { "epoch": 2.353390153924437, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.873246192932129, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8641247153282166, "num_tokens": 705921944.0, "step": 18500 }, { "epoch": 2.353517364203028, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.81482982635498, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8739612698554993, "num_tokens": 705967999.0, "step": 18501 }, { "epoch": 2.353644574481618, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.820661544799805, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8659427762031555, "num_tokens": 706003370.0, "step": 18502 }, { "epoch": 2.353771784760209, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.831746101379395, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8629668951034546, "num_tokens": 706036551.0, "step": 18503 }, { "epoch": 2.353898995038799, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.861428260803223, "learning_rate": 1e-06, "loss": 0.3443, "mean_token_accuracy": 0.8915778398513794, "num_tokens": 706075965.0, "step": 18504 }, { "epoch": 2.3540262053173895, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.832508087158203, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8617585897445679, "num_tokens": 706117943.0, "step": 18505 }, { "epoch": 2.35415341559598, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.85958194732666, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8770666122436523, "num_tokens": 706157183.0, "step": 18506 }, { "epoch": 2.3542806258745705, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.8340425491333, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8816942572593689, "num_tokens": 706196878.0, "step": 18507 }, { "epoch": 2.354407836153161, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.856871604919434, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.860639214515686, "num_tokens": 706231102.0, "step": 18508 }, { "epoch": 2.3545350464317516, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.815058708190918, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8811565637588501, "num_tokens": 706273372.0, "step": 18509 }, { "epoch": 2.354662256710342, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.870271682739258, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8668591976165771, "num_tokens": 706308106.0, "step": 18510 }, { "epoch": 2.3547894669889327, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.810859680175781, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8808949589729309, "num_tokens": 706344267.0, "step": 18511 }, { "epoch": 2.354916677267523, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.82125186920166, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8916109204292297, "num_tokens": 706385019.0, "step": 18512 }, { "epoch": 2.3550438875461137, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.839319229125977, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8726059198379517, "num_tokens": 706424913.0, "step": 18513 }, { "epoch": 2.3551710978247042, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.83365535736084, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8736549615859985, "num_tokens": 706466064.0, "step": 18514 }, { "epoch": 2.3552983081032948, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.824277877807617, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8802170753479004, "num_tokens": 706504572.0, "step": 18515 }, { "epoch": 2.3554255183818853, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.874834060668945, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8676284551620483, "num_tokens": 706543901.0, "step": 18516 }, { "epoch": 2.355552728660476, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.84826946258545, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8669165372848511, "num_tokens": 706579095.0, "step": 18517 }, { "epoch": 2.3556799389390664, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.885201454162598, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.866023063659668, "num_tokens": 706614078.0, "step": 18518 }, { "epoch": 2.355807149217657, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.818312644958496, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8809793591499329, "num_tokens": 706652596.0, "step": 18519 }, { "epoch": 2.3559343594962474, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.849289894104004, "learning_rate": 1e-06, "loss": 0.3285, "mean_token_accuracy": 0.8968492746353149, "num_tokens": 706691297.0, "step": 18520 }, { "epoch": 2.356061569774838, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.846878051757812, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8629432916641235, "num_tokens": 706727344.0, "step": 18521 }, { "epoch": 2.3561887800534285, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.825725555419922, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8770340085029602, "num_tokens": 706770411.0, "step": 18522 }, { "epoch": 2.356315990332019, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.893863677978516, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8765047192573547, "num_tokens": 706816215.0, "step": 18523 }, { "epoch": 2.3564432006106095, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.841115951538086, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8681445717811584, "num_tokens": 706857400.0, "step": 18524 }, { "epoch": 2.3565704108891996, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.903705596923828, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8663389682769775, "num_tokens": 706892715.0, "step": 18525 }, { "epoch": 2.3566976211677906, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.875969886779785, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8750283718109131, "num_tokens": 706934373.0, "step": 18526 }, { "epoch": 2.3568248314463807, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.906116485595703, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8751181364059448, "num_tokens": 706976885.0, "step": 18527 }, { "epoch": 2.356952041724971, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.81095027923584, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.87940514087677, "num_tokens": 707016957.0, "step": 18528 }, { "epoch": 2.3570792520035617, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.912328720092773, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8831917643547058, "num_tokens": 707053687.0, "step": 18529 }, { "epoch": 2.3572064622821522, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.863723754882812, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8673259019851685, "num_tokens": 707087327.0, "step": 18530 }, { "epoch": 2.3573336725607428, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.84684944152832, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8858710527420044, "num_tokens": 707124479.0, "step": 18531 }, { "epoch": 2.3574608828393333, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.88416576385498, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8782564401626587, "num_tokens": 707162452.0, "step": 18532 }, { "epoch": 2.357588093117924, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.8844575881958, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.884086012840271, "num_tokens": 707196765.0, "step": 18533 }, { "epoch": 2.3577153033965144, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.96577262878418, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8792715072631836, "num_tokens": 707230151.0, "step": 18534 }, { "epoch": 2.357842513675105, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.900535583496094, "learning_rate": 1e-06, "loss": 0.4514, "mean_token_accuracy": 0.8565965890884399, "num_tokens": 707264373.0, "step": 18535 }, { "epoch": 2.3579697239536954, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.865797996520996, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8686435222625732, "num_tokens": 707304844.0, "step": 18536 }, { "epoch": 2.358096934232286, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.856468200683594, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8756231069564819, "num_tokens": 707347081.0, "step": 18537 }, { "epoch": 2.3582241445108765, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.843289375305176, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8676609992980957, "num_tokens": 707388243.0, "step": 18538 }, { "epoch": 2.358351354789467, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.858949661254883, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8603000044822693, "num_tokens": 707427109.0, "step": 18539 }, { "epoch": 2.3584785650680575, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.923386573791504, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8757843375205994, "num_tokens": 707462502.0, "step": 18540 }, { "epoch": 2.358605775346648, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.844685554504395, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8679330945014954, "num_tokens": 707502861.0, "step": 18541 }, { "epoch": 2.3587329856252386, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.875544548034668, "learning_rate": 1e-06, "loss": 0.4615, "mean_token_accuracy": 0.8546155095100403, "num_tokens": 707535868.0, "step": 18542 }, { "epoch": 2.358860195903829, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.911140441894531, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.876652717590332, "num_tokens": 707566128.0, "step": 18543 }, { "epoch": 2.3589874061824196, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.84090805053711, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8800694942474365, "num_tokens": 707605113.0, "step": 18544 }, { "epoch": 2.35911461646101, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.836773872375488, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8724989891052246, "num_tokens": 707647968.0, "step": 18545 }, { "epoch": 2.3592418267396007, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.849350929260254, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8815522193908691, "num_tokens": 707682183.0, "step": 18546 }, { "epoch": 2.3593690370181912, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.753255844116211, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8824245929718018, "num_tokens": 707720938.0, "step": 18547 }, { "epoch": 2.3594962472967818, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.873177528381348, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8621950149536133, "num_tokens": 707762624.0, "step": 18548 }, { "epoch": 2.3596234575753723, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.791725158691406, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8666550517082214, "num_tokens": 707798406.0, "step": 18549 }, { "epoch": 2.3597506678539624, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.851465225219727, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8756816387176514, "num_tokens": 707835935.0, "step": 18550 }, { "epoch": 2.3598778781325533, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.793485641479492, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.885358452796936, "num_tokens": 707872971.0, "step": 18551 }, { "epoch": 2.3600050884111434, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.923687934875488, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8851624727249146, "num_tokens": 707909798.0, "step": 18552 }, { "epoch": 2.360132298689734, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.749980926513672, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8768929243087769, "num_tokens": 707951130.0, "step": 18553 }, { "epoch": 2.3602595089683245, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.814102172851562, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8617094159126282, "num_tokens": 707989502.0, "step": 18554 }, { "epoch": 2.360386719246915, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.862658500671387, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8732653260231018, "num_tokens": 708023521.0, "step": 18555 }, { "epoch": 2.3605139295255055, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.863116264343262, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8804683685302734, "num_tokens": 708057594.0, "step": 18556 }, { "epoch": 2.360641139804096, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.838343620300293, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8716330528259277, "num_tokens": 708101393.0, "step": 18557 }, { "epoch": 2.3607683500826866, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.814650535583496, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8846511840820312, "num_tokens": 708137141.0, "step": 18558 }, { "epoch": 2.360895560361277, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.871071815490723, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8791910409927368, "num_tokens": 708173497.0, "step": 18559 }, { "epoch": 2.3610227706398677, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.893036842346191, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8734747171401978, "num_tokens": 708209412.0, "step": 18560 }, { "epoch": 2.361149980918458, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.828471183776855, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8812124729156494, "num_tokens": 708239675.0, "step": 18561 }, { "epoch": 2.3612771911970487, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.88198471069336, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8679762482643127, "num_tokens": 708274918.0, "step": 18562 }, { "epoch": 2.3614044014756392, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.810176849365234, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.859074592590332, "num_tokens": 708313645.0, "step": 18563 }, { "epoch": 2.3615316117542298, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.84002685546875, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8749771118164062, "num_tokens": 708357985.0, "step": 18564 }, { "epoch": 2.3616588220328203, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.832242965698242, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.8535768389701843, "num_tokens": 708393606.0, "step": 18565 }, { "epoch": 2.361786032311411, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.860581398010254, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8789300918579102, "num_tokens": 708428941.0, "step": 18566 }, { "epoch": 2.3619132425900013, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.820420265197754, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8779317736625671, "num_tokens": 708467847.0, "step": 18567 }, { "epoch": 2.362040452868592, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.796521186828613, "learning_rate": 1e-06, "loss": 0.3418, "mean_token_accuracy": 0.8899397253990173, "num_tokens": 708503212.0, "step": 18568 }, { "epoch": 2.3621676631471824, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.865466117858887, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8746641874313354, "num_tokens": 708542897.0, "step": 18569 }, { "epoch": 2.362294873425773, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.860873222351074, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8619155883789062, "num_tokens": 708574725.0, "step": 18570 }, { "epoch": 2.3624220837043635, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.848052024841309, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8841515779495239, "num_tokens": 708613557.0, "step": 18571 }, { "epoch": 2.362549293982954, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.80971622467041, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8659946918487549, "num_tokens": 708651872.0, "step": 18572 }, { "epoch": 2.3626765042615445, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.846028327941895, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8616112470626831, "num_tokens": 708688654.0, "step": 18573 }, { "epoch": 2.362803714540135, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.807941436767578, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8803615570068359, "num_tokens": 708723285.0, "step": 18574 }, { "epoch": 2.362930924818725, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.915772438049316, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.883145809173584, "num_tokens": 708760617.0, "step": 18575 }, { "epoch": 2.363058135097316, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.826484680175781, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8782985210418701, "num_tokens": 708790792.0, "step": 18576 }, { "epoch": 2.363185345375906, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.905773162841797, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8648560047149658, "num_tokens": 708829935.0, "step": 18577 }, { "epoch": 2.3633125556544967, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.822314262390137, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8695022463798523, "num_tokens": 708870578.0, "step": 18578 }, { "epoch": 2.3634397659330872, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.874226570129395, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8718703985214233, "num_tokens": 708907044.0, "step": 18579 }, { "epoch": 2.3635669762116778, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.875080108642578, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8560848236083984, "num_tokens": 708947443.0, "step": 18580 }, { "epoch": 2.3636941864902683, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.872743606567383, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8759876489639282, "num_tokens": 708983174.0, "step": 18581 }, { "epoch": 2.363821396768859, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.826135635375977, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.877086877822876, "num_tokens": 709021950.0, "step": 18582 }, { "epoch": 2.3639486070474494, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.848485946655273, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8797631859779358, "num_tokens": 709057236.0, "step": 18583 }, { "epoch": 2.36407581732604, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.889052391052246, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8749728202819824, "num_tokens": 709093376.0, "step": 18584 }, { "epoch": 2.3642030276046304, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.883130073547363, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8692928552627563, "num_tokens": 709129502.0, "step": 18585 }, { "epoch": 2.364330237883221, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.871049880981445, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8741350769996643, "num_tokens": 709162538.0, "step": 18586 }, { "epoch": 2.3644574481618115, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.83909797668457, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8636752367019653, "num_tokens": 709201690.0, "step": 18587 }, { "epoch": 2.364584658440402, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.899944305419922, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8676466941833496, "num_tokens": 709240593.0, "step": 18588 }, { "epoch": 2.3647118687189925, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.850611686706543, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8845869898796082, "num_tokens": 709278192.0, "step": 18589 }, { "epoch": 2.364839078997583, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.826812744140625, "learning_rate": 1e-06, "loss": 0.3365, "mean_token_accuracy": 0.890678346157074, "num_tokens": 709311797.0, "step": 18590 }, { "epoch": 2.3649662892761736, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.903549194335938, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8717933297157288, "num_tokens": 709350379.0, "step": 18591 }, { "epoch": 2.365093499554764, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.832496643066406, "learning_rate": 1e-06, "loss": 0.3234, "mean_token_accuracy": 0.8983160257339478, "num_tokens": 709389030.0, "step": 18592 }, { "epoch": 2.3652207098333546, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.851304054260254, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8655441999435425, "num_tokens": 709433692.0, "step": 18593 }, { "epoch": 2.365347920111945, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.865769386291504, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8736518621444702, "num_tokens": 709470773.0, "step": 18594 }, { "epoch": 2.3654751303905357, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.877842903137207, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8691447973251343, "num_tokens": 709506764.0, "step": 18595 }, { "epoch": 2.3656023406691262, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.897441864013672, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8728418946266174, "num_tokens": 709546931.0, "step": 18596 }, { "epoch": 2.3657295509477168, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.870676040649414, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.874682605266571, "num_tokens": 709582842.0, "step": 18597 }, { "epoch": 2.365856761226307, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.81689167022705, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8734645843505859, "num_tokens": 709621574.0, "step": 18598 }, { "epoch": 2.365983971504898, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.85175895690918, "learning_rate": 1e-06, "loss": 0.3431, "mean_token_accuracy": 0.8914259076118469, "num_tokens": 709662437.0, "step": 18599 }, { "epoch": 2.366111181783488, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.850080490112305, "learning_rate": 1e-06, "loss": 0.3542, "mean_token_accuracy": 0.887336254119873, "num_tokens": 709701318.0, "step": 18600 }, { "epoch": 2.3662383920620784, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.909004211425781, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8738224506378174, "num_tokens": 709745156.0, "step": 18601 }, { "epoch": 2.366365602340669, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.843036651611328, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8695726990699768, "num_tokens": 709785849.0, "step": 18602 }, { "epoch": 2.3664928126192595, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.796357154846191, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8777674436569214, "num_tokens": 709830038.0, "step": 18603 }, { "epoch": 2.36662002289785, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.870018005371094, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8720865249633789, "num_tokens": 709871174.0, "step": 18604 }, { "epoch": 2.3667472331764405, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.820014953613281, "learning_rate": 1e-06, "loss": 0.3327, "mean_token_accuracy": 0.8949846029281616, "num_tokens": 709905921.0, "step": 18605 }, { "epoch": 2.366874443455031, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.906039237976074, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8893829584121704, "num_tokens": 709939856.0, "step": 18606 }, { "epoch": 2.3670016537336216, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.855813980102539, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8859928250312805, "num_tokens": 709978805.0, "step": 18607 }, { "epoch": 2.367128864012212, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.834029197692871, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.887872576713562, "num_tokens": 710015015.0, "step": 18608 }, { "epoch": 2.3672560742908026, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.92708683013916, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.865231990814209, "num_tokens": 710054887.0, "step": 18609 }, { "epoch": 2.367383284569393, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.899810791015625, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8745779991149902, "num_tokens": 710097012.0, "step": 18610 }, { "epoch": 2.3675104948479837, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.921685218811035, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8582760095596313, "num_tokens": 710130912.0, "step": 18611 }, { "epoch": 2.3676377051265742, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.872682571411133, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8701618313789368, "num_tokens": 710168880.0, "step": 18612 }, { "epoch": 2.3677649154051648, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.901432991027832, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.883143424987793, "num_tokens": 710205619.0, "step": 18613 }, { "epoch": 2.3678921256837553, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.92856502532959, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8805363774299622, "num_tokens": 710239475.0, "step": 18614 }, { "epoch": 2.368019335962346, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.922272682189941, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8764713406562805, "num_tokens": 710276533.0, "step": 18615 }, { "epoch": 2.3681465462409363, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.902153015136719, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.8898467421531677, "num_tokens": 710316731.0, "step": 18616 }, { "epoch": 2.368273756519527, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.938164710998535, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8709125518798828, "num_tokens": 710352310.0, "step": 18617 }, { "epoch": 2.3684009667981174, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.92888355255127, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.870276689529419, "num_tokens": 710400203.0, "step": 18618 }, { "epoch": 2.368528177076708, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.88978385925293, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8802307844161987, "num_tokens": 710433054.0, "step": 18619 }, { "epoch": 2.3686553873552985, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.876471519470215, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8786035776138306, "num_tokens": 710466879.0, "step": 18620 }, { "epoch": 2.368782597633889, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.894063949584961, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.868207573890686, "num_tokens": 710502808.0, "step": 18621 }, { "epoch": 2.3689098079124795, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.866546630859375, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8634257316589355, "num_tokens": 710546190.0, "step": 18622 }, { "epoch": 2.3690370181910696, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.856674194335938, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8767598271369934, "num_tokens": 710585556.0, "step": 18623 }, { "epoch": 2.3691642284696606, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.951513290405273, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8772799372673035, "num_tokens": 710621366.0, "step": 18624 }, { "epoch": 2.3692914387482507, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.831342697143555, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8782264590263367, "num_tokens": 710660217.0, "step": 18625 }, { "epoch": 2.369418649026841, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.945037841796875, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8711583018302917, "num_tokens": 710695798.0, "step": 18626 }, { "epoch": 2.3695458593054317, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.90064525604248, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8576666712760925, "num_tokens": 710732739.0, "step": 18627 }, { "epoch": 2.3696730695840222, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.970619201660156, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8667742013931274, "num_tokens": 710769651.0, "step": 18628 }, { "epoch": 2.3698002798626128, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.858899116516113, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8727178573608398, "num_tokens": 710809487.0, "step": 18629 }, { "epoch": 2.3699274901412033, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.91804313659668, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8767735362052917, "num_tokens": 710849152.0, "step": 18630 }, { "epoch": 2.370054700419794, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.90776538848877, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8612900376319885, "num_tokens": 710890933.0, "step": 18631 }, { "epoch": 2.3701819106983844, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.843684196472168, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8776580095291138, "num_tokens": 710928262.0, "step": 18632 }, { "epoch": 2.370309120976975, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.973173141479492, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8797046542167664, "num_tokens": 710962969.0, "step": 18633 }, { "epoch": 2.3704363312555654, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.973532676696777, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8698424696922302, "num_tokens": 710998717.0, "step": 18634 }, { "epoch": 2.370563541534156, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.893014907836914, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.860299825668335, "num_tokens": 711032819.0, "step": 18635 }, { "epoch": 2.3706907518127465, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.92597770690918, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.865515947341919, "num_tokens": 711069395.0, "step": 18636 }, { "epoch": 2.370817962091337, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.815105438232422, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8766953945159912, "num_tokens": 711105754.0, "step": 18637 }, { "epoch": 2.3709451723699275, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.918180465698242, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8689424991607666, "num_tokens": 711145291.0, "step": 18638 }, { "epoch": 2.371072382648518, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.863582611083984, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.875594973564148, "num_tokens": 711183431.0, "step": 18639 }, { "epoch": 2.3711995929271086, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.859832763671875, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8591228127479553, "num_tokens": 711229754.0, "step": 18640 }, { "epoch": 2.371326803205699, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.882616996765137, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8753688335418701, "num_tokens": 711269071.0, "step": 18641 }, { "epoch": 2.3714540134842896, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.917252540588379, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8710679411888123, "num_tokens": 711312036.0, "step": 18642 }, { "epoch": 2.37158122376288, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.881505012512207, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8812218904495239, "num_tokens": 711347945.0, "step": 18643 }, { "epoch": 2.3717084340414707, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.956947326660156, "learning_rate": 1e-06, "loss": 0.4681, "mean_token_accuracy": 0.853517472743988, "num_tokens": 711388433.0, "step": 18644 }, { "epoch": 2.371835644320061, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.877790451049805, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8785495758056641, "num_tokens": 711425821.0, "step": 18645 }, { "epoch": 2.3719628545986517, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.915947914123535, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8684901595115662, "num_tokens": 711462725.0, "step": 18646 }, { "epoch": 2.3720900648772423, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.875364303588867, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8872514963150024, "num_tokens": 711499987.0, "step": 18647 }, { "epoch": 2.3722172751558324, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.892243385314941, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8595936298370361, "num_tokens": 711543658.0, "step": 18648 }, { "epoch": 2.3723444854344233, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.842275619506836, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8851933479309082, "num_tokens": 711583875.0, "step": 18649 }, { "epoch": 2.3724716957130134, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.862516403198242, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8570449352264404, "num_tokens": 711625543.0, "step": 18650 }, { "epoch": 2.372598905991604, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.871651649475098, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8766390085220337, "num_tokens": 711660528.0, "step": 18651 }, { "epoch": 2.3727261162701945, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.789254188537598, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8891452550888062, "num_tokens": 711695646.0, "step": 18652 }, { "epoch": 2.372853326548785, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.854422569274902, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8589416742324829, "num_tokens": 711734691.0, "step": 18653 }, { "epoch": 2.3729805368273755, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.8569974899292, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8761138916015625, "num_tokens": 711771518.0, "step": 18654 }, { "epoch": 2.373107747105966, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.91637134552002, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8635563254356384, "num_tokens": 711808968.0, "step": 18655 }, { "epoch": 2.3732349573845566, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.836922645568848, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8815237283706665, "num_tokens": 711842944.0, "step": 18656 }, { "epoch": 2.373362167663147, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.8865327835083, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.870824933052063, "num_tokens": 711882740.0, "step": 18657 }, { "epoch": 2.3734893779417376, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.888320922851562, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.867011308670044, "num_tokens": 711925362.0, "step": 18658 }, { "epoch": 2.373616588220328, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.886432647705078, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8780588507652283, "num_tokens": 711971285.0, "step": 18659 }, { "epoch": 2.3737437984989187, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.840229988098145, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8695197105407715, "num_tokens": 712013306.0, "step": 18660 }, { "epoch": 2.3738710087775092, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.962796211242676, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8743475675582886, "num_tokens": 712050837.0, "step": 18661 }, { "epoch": 2.3739982190560998, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.868462562561035, "learning_rate": 1e-06, "loss": 0.3513, "mean_token_accuracy": 0.8907076716423035, "num_tokens": 712085140.0, "step": 18662 }, { "epoch": 2.3741254293346903, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.863029479980469, "learning_rate": 1e-06, "loss": 0.357, "mean_token_accuracy": 0.8865624070167542, "num_tokens": 712122495.0, "step": 18663 }, { "epoch": 2.374252639613281, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.906864166259766, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8672609329223633, "num_tokens": 712161312.0, "step": 18664 }, { "epoch": 2.3743798498918713, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.915210723876953, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8732317686080933, "num_tokens": 712197275.0, "step": 18665 }, { "epoch": 2.374507060170462, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.870625495910645, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8626086711883545, "num_tokens": 712231309.0, "step": 18666 }, { "epoch": 2.3746342704490524, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.884930610656738, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.869198739528656, "num_tokens": 712264864.0, "step": 18667 }, { "epoch": 2.374761480727643, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.898259162902832, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.873806893825531, "num_tokens": 712303564.0, "step": 18668 }, { "epoch": 2.3748886910062335, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.883513450622559, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.884615421295166, "num_tokens": 712341872.0, "step": 18669 }, { "epoch": 2.375015901284824, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.90526008605957, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8626497387886047, "num_tokens": 712379835.0, "step": 18670 }, { "epoch": 2.3751431115634145, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.882790565490723, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8698183298110962, "num_tokens": 712417462.0, "step": 18671 }, { "epoch": 2.375270321842005, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.840314865112305, "learning_rate": 1e-06, "loss": 0.4871, "mean_token_accuracy": 0.8499218821525574, "num_tokens": 712461581.0, "step": 18672 }, { "epoch": 2.375397532120595, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.863218307495117, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8724706768989563, "num_tokens": 712499615.0, "step": 18673 }, { "epoch": 2.375524742399186, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.902573585510254, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8842183351516724, "num_tokens": 712533264.0, "step": 18674 }, { "epoch": 2.375651952677776, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.828022003173828, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8662613034248352, "num_tokens": 712571496.0, "step": 18675 }, { "epoch": 2.3757791629563667, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.924357414245605, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8809703588485718, "num_tokens": 712606510.0, "step": 18676 }, { "epoch": 2.3759063732349572, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.848544120788574, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8829048275947571, "num_tokens": 712646218.0, "step": 18677 }, { "epoch": 2.3760335835135478, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.943793296813965, "learning_rate": 1e-06, "loss": 0.4976, "mean_token_accuracy": 0.8428903818130493, "num_tokens": 712681905.0, "step": 18678 }, { "epoch": 2.3761607937921383, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.848518371582031, "learning_rate": 1e-06, "loss": 0.3492, "mean_token_accuracy": 0.8863909244537354, "num_tokens": 712719916.0, "step": 18679 }, { "epoch": 2.376288004070729, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.938036918640137, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8739192485809326, "num_tokens": 712760849.0, "step": 18680 }, { "epoch": 2.3764152143493193, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.843643188476562, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8848595023155212, "num_tokens": 712801199.0, "step": 18681 }, { "epoch": 2.37654242462791, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.913233757019043, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8683769702911377, "num_tokens": 712843471.0, "step": 18682 }, { "epoch": 2.3766696349065004, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.841595649719238, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.874441385269165, "num_tokens": 712880649.0, "step": 18683 }, { "epoch": 2.376796845185091, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.79379940032959, "learning_rate": 1e-06, "loss": 0.4642, "mean_token_accuracy": 0.8499622941017151, "num_tokens": 712920574.0, "step": 18684 }, { "epoch": 2.3769240554636815, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.936239242553711, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8819394707679749, "num_tokens": 712962544.0, "step": 18685 }, { "epoch": 2.377051265742272, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.825275421142578, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8775887489318848, "num_tokens": 712999138.0, "step": 18686 }, { "epoch": 2.3771784760208625, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.934826850891113, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8751575946807861, "num_tokens": 713041799.0, "step": 18687 }, { "epoch": 2.377305686299453, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.820003509521484, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8660854697227478, "num_tokens": 713088248.0, "step": 18688 }, { "epoch": 2.3774328965780436, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.886305809020996, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8828104138374329, "num_tokens": 713127239.0, "step": 18689 }, { "epoch": 2.377560106856634, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.918241500854492, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8766541481018066, "num_tokens": 713168389.0, "step": 18690 }, { "epoch": 2.3776873171352246, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.875778198242188, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8670591115951538, "num_tokens": 713208137.0, "step": 18691 }, { "epoch": 2.377814527413815, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.9085693359375, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8644740581512451, "num_tokens": 713249344.0, "step": 18692 }, { "epoch": 2.3779417376924057, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.919642448425293, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8696267604827881, "num_tokens": 713291376.0, "step": 18693 }, { "epoch": 2.378068947970996, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.840360641479492, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8722670078277588, "num_tokens": 713329925.0, "step": 18694 }, { "epoch": 2.3781961582495867, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.905891418457031, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8754736185073853, "num_tokens": 713372509.0, "step": 18695 }, { "epoch": 2.378323368528177, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.89969539642334, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.868131160736084, "num_tokens": 713410055.0, "step": 18696 }, { "epoch": 2.378450578806768, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.885079383850098, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8720076084136963, "num_tokens": 713453958.0, "step": 18697 }, { "epoch": 2.378577789085358, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.860962867736816, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8732321858406067, "num_tokens": 713485683.0, "step": 18698 }, { "epoch": 2.3787049993639484, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.930865287780762, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.875178873538971, "num_tokens": 713528565.0, "step": 18699 }, { "epoch": 2.378832209642539, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.900091171264648, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8703296184539795, "num_tokens": 713560932.0, "step": 18700 }, { "epoch": 2.3789594199211295, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.932144165039062, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8707633018493652, "num_tokens": 713592891.0, "step": 18701 }, { "epoch": 2.37908663019972, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.893163681030273, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8712863922119141, "num_tokens": 713631285.0, "step": 18702 }, { "epoch": 2.3792138404783105, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.86284351348877, "learning_rate": 1e-06, "loss": 0.4655, "mean_token_accuracy": 0.8558059930801392, "num_tokens": 713675143.0, "step": 18703 }, { "epoch": 2.379341050756901, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.84417724609375, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8774802684783936, "num_tokens": 713717556.0, "step": 18704 }, { "epoch": 2.3794682610354916, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.882230758666992, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8658950328826904, "num_tokens": 713763515.0, "step": 18705 }, { "epoch": 2.379595471314082, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.890042304992676, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8783513307571411, "num_tokens": 713801064.0, "step": 18706 }, { "epoch": 2.3797226815926726, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.813112258911133, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8704113364219666, "num_tokens": 713841662.0, "step": 18707 }, { "epoch": 2.379849891871263, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.866148948669434, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8760680556297302, "num_tokens": 713883646.0, "step": 18708 }, { "epoch": 2.3799771021498537, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.841690063476562, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.8932408690452576, "num_tokens": 713925365.0, "step": 18709 }, { "epoch": 2.3801043124284442, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.85412311553955, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8703069090843201, "num_tokens": 713961142.0, "step": 18710 }, { "epoch": 2.3802315227070348, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.868882179260254, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8812329769134521, "num_tokens": 714001652.0, "step": 18711 }, { "epoch": 2.3803587329856253, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.840121269226074, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8757112622261047, "num_tokens": 714043023.0, "step": 18712 }, { "epoch": 2.380485943264216, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.909945487976074, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8737539649009705, "num_tokens": 714080389.0, "step": 18713 }, { "epoch": 2.3806131535428063, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.899227142333984, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.877575159072876, "num_tokens": 714117695.0, "step": 18714 }, { "epoch": 2.380740363821397, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.894794464111328, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8803704977035522, "num_tokens": 714154204.0, "step": 18715 }, { "epoch": 2.3808675740999874, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.892284393310547, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8570189476013184, "num_tokens": 714194857.0, "step": 18716 }, { "epoch": 2.380994784378578, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.893997192382812, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8849003314971924, "num_tokens": 714228034.0, "step": 18717 }, { "epoch": 2.3811219946571685, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.891884803771973, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8796249628067017, "num_tokens": 714269688.0, "step": 18718 }, { "epoch": 2.381249204935759, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.846579551696777, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8686370849609375, "num_tokens": 714309039.0, "step": 18719 }, { "epoch": 2.3813764152143495, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.953862190246582, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8792977333068848, "num_tokens": 714347453.0, "step": 18720 }, { "epoch": 2.3815036254929396, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.840311050415039, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8712782859802246, "num_tokens": 714385850.0, "step": 18721 }, { "epoch": 2.3816308357715306, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.873343467712402, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.868942379951477, "num_tokens": 714426735.0, "step": 18722 }, { "epoch": 2.3817580460501206, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.926162719726562, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8853588700294495, "num_tokens": 714465279.0, "step": 18723 }, { "epoch": 2.381885256328711, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.839376449584961, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8631614446640015, "num_tokens": 714505703.0, "step": 18724 }, { "epoch": 2.3820124666073017, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.99772834777832, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8674520254135132, "num_tokens": 714543642.0, "step": 18725 }, { "epoch": 2.3821396768858922, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.824726104736328, "learning_rate": 1e-06, "loss": 0.3494, "mean_token_accuracy": 0.8901629447937012, "num_tokens": 714583533.0, "step": 18726 }, { "epoch": 2.3822668871644828, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.926336288452148, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8787733912467957, "num_tokens": 714625577.0, "step": 18727 }, { "epoch": 2.3823940974430733, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.856356620788574, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8704813718795776, "num_tokens": 714664987.0, "step": 18728 }, { "epoch": 2.382521307721664, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.958362579345703, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8670378923416138, "num_tokens": 714705881.0, "step": 18729 }, { "epoch": 2.3826485180002543, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.899563789367676, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8642125129699707, "num_tokens": 714746012.0, "step": 18730 }, { "epoch": 2.382775728278845, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.931448936462402, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8753311634063721, "num_tokens": 714790001.0, "step": 18731 }, { "epoch": 2.3829029385574354, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.864819526672363, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8804524540901184, "num_tokens": 714824515.0, "step": 18732 }, { "epoch": 2.383030148836026, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.965899467468262, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.866302490234375, "num_tokens": 714860841.0, "step": 18733 }, { "epoch": 2.3831573591146165, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.853476524353027, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8763340711593628, "num_tokens": 714896373.0, "step": 18734 }, { "epoch": 2.383284569393207, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.912666320800781, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8774881362915039, "num_tokens": 714926684.0, "step": 18735 }, { "epoch": 2.3834117796717975, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.892189025878906, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8876569271087646, "num_tokens": 714961057.0, "step": 18736 }, { "epoch": 2.383538989950388, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.895939826965332, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8794088959693909, "num_tokens": 715001215.0, "step": 18737 }, { "epoch": 2.3836662002289786, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.963680267333984, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8544855117797852, "num_tokens": 715038252.0, "step": 18738 }, { "epoch": 2.383793410507569, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.869606018066406, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.872603714466095, "num_tokens": 715078320.0, "step": 18739 }, { "epoch": 2.3839206207861596, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.924474716186523, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.868541955947876, "num_tokens": 715115628.0, "step": 18740 }, { "epoch": 2.38404783106475, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.908025741577148, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8826545476913452, "num_tokens": 715153294.0, "step": 18741 }, { "epoch": 2.3841750413433407, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.93415641784668, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8746663928031921, "num_tokens": 715193736.0, "step": 18742 }, { "epoch": 2.384302251621931, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.921266555786133, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8752501010894775, "num_tokens": 715234539.0, "step": 18743 }, { "epoch": 2.3844294619005217, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.93056583404541, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8859519958496094, "num_tokens": 715271567.0, "step": 18744 }, { "epoch": 2.3845566721791123, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.882946968078613, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8855562806129456, "num_tokens": 715314850.0, "step": 18745 }, { "epoch": 2.3846838824577024, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.90457820892334, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8805124759674072, "num_tokens": 715353353.0, "step": 18746 }, { "epoch": 2.3848110927362933, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.946001052856445, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8850566148757935, "num_tokens": 715387843.0, "step": 18747 }, { "epoch": 2.3849383030148834, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.852302551269531, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8615200519561768, "num_tokens": 715428037.0, "step": 18748 }, { "epoch": 2.385065513293474, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.883211135864258, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.887279748916626, "num_tokens": 715464403.0, "step": 18749 }, { "epoch": 2.3851927235720645, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.923638343811035, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8839004039764404, "num_tokens": 715501820.0, "step": 18750 }, { "epoch": 2.385319933850655, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.876731872558594, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.883169949054718, "num_tokens": 715543287.0, "step": 18751 }, { "epoch": 2.3854471441292455, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.899409294128418, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8832898736000061, "num_tokens": 715578189.0, "step": 18752 }, { "epoch": 2.385574354407836, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.932358741760254, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8647058606147766, "num_tokens": 715616447.0, "step": 18753 }, { "epoch": 2.3857015646864266, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.950276374816895, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8782746195793152, "num_tokens": 715651692.0, "step": 18754 }, { "epoch": 2.385828774965017, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.900604248046875, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8670819997787476, "num_tokens": 715687733.0, "step": 18755 }, { "epoch": 2.3859559852436076, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.858473777770996, "learning_rate": 1e-06, "loss": 0.4552, "mean_token_accuracy": 0.8530710935592651, "num_tokens": 715727481.0, "step": 18756 }, { "epoch": 2.386083195522198, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.912878036499023, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8569692969322205, "num_tokens": 715767549.0, "step": 18757 }, { "epoch": 2.3862104058007887, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.919721603393555, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8678961992263794, "num_tokens": 715807251.0, "step": 18758 }, { "epoch": 2.386337616079379, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.905844688415527, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8586151599884033, "num_tokens": 715846893.0, "step": 18759 }, { "epoch": 2.3864648263579697, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.888066291809082, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8618254661560059, "num_tokens": 715883611.0, "step": 18760 }, { "epoch": 2.3865920366365603, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.890345573425293, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8756524324417114, "num_tokens": 715915598.0, "step": 18761 }, { "epoch": 2.386719246915151, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.900492668151855, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8807766437530518, "num_tokens": 715956647.0, "step": 18762 }, { "epoch": 2.3868464571937413, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.962289810180664, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8809316158294678, "num_tokens": 715996627.0, "step": 18763 }, { "epoch": 2.386973667472332, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.862894058227539, "learning_rate": 1e-06, "loss": 0.3322, "mean_token_accuracy": 0.8937220573425293, "num_tokens": 716027823.0, "step": 18764 }, { "epoch": 2.3871008777509224, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.891735076904297, "learning_rate": 1e-06, "loss": 0.4462, "mean_token_accuracy": 0.8609756231307983, "num_tokens": 716067430.0, "step": 18765 }, { "epoch": 2.387228088029513, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.925829887390137, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8665131330490112, "num_tokens": 716109436.0, "step": 18766 }, { "epoch": 2.3873552983081034, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.900160789489746, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.869868814945221, "num_tokens": 716148157.0, "step": 18767 }, { "epoch": 2.387482508586694, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.9248628616333, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8765105605125427, "num_tokens": 716184662.0, "step": 18768 }, { "epoch": 2.3876097188652845, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.960867881774902, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8666305541992188, "num_tokens": 716227382.0, "step": 18769 }, { "epoch": 2.387736929143875, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.864344596862793, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8769229650497437, "num_tokens": 716262236.0, "step": 18770 }, { "epoch": 2.387864139422465, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.936480522155762, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8814554810523987, "num_tokens": 716295541.0, "step": 18771 }, { "epoch": 2.387991349701056, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.878705978393555, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.877414345741272, "num_tokens": 716331796.0, "step": 18772 }, { "epoch": 2.388118559979646, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.848750114440918, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.86451655626297, "num_tokens": 716369539.0, "step": 18773 }, { "epoch": 2.3882457702582367, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.87488842010498, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8909934759140015, "num_tokens": 716409806.0, "step": 18774 }, { "epoch": 2.3883729805368272, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.861559867858887, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8659274578094482, "num_tokens": 716448191.0, "step": 18775 }, { "epoch": 2.3885001908154178, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.840372085571289, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8748332262039185, "num_tokens": 716486522.0, "step": 18776 }, { "epoch": 2.3886274010940083, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.90613842010498, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8757553100585938, "num_tokens": 716519383.0, "step": 18777 }, { "epoch": 2.388754611372599, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.830838203430176, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.87575364112854, "num_tokens": 716556136.0, "step": 18778 }, { "epoch": 2.3888818216511893, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.904780387878418, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8673412799835205, "num_tokens": 716592346.0, "step": 18779 }, { "epoch": 2.38900903192978, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.874943733215332, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.86733478307724, "num_tokens": 716633996.0, "step": 18780 }, { "epoch": 2.3891362422083704, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.910469055175781, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8690648674964905, "num_tokens": 716669638.0, "step": 18781 }, { "epoch": 2.389263452486961, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.86661434173584, "learning_rate": 1e-06, "loss": 0.419, "mean_token_accuracy": 0.8651270270347595, "num_tokens": 716711194.0, "step": 18782 }, { "epoch": 2.3893906627655515, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.861209869384766, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8801881074905396, "num_tokens": 716753402.0, "step": 18783 }, { "epoch": 2.389517873044142, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.909377098083496, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8754730820655823, "num_tokens": 716794638.0, "step": 18784 }, { "epoch": 2.3896450833227325, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.856094360351562, "learning_rate": 1e-06, "loss": 0.4765, "mean_token_accuracy": 0.8454892635345459, "num_tokens": 716834003.0, "step": 18785 }, { "epoch": 2.389772293601323, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.871938705444336, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.879845917224884, "num_tokens": 716867349.0, "step": 18786 }, { "epoch": 2.3898995038799136, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.949104309082031, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8727967739105225, "num_tokens": 716904271.0, "step": 18787 }, { "epoch": 2.390026714158504, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.948885917663574, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8770376443862915, "num_tokens": 716936165.0, "step": 18788 }, { "epoch": 2.3901539244370946, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.852465629577637, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8728458881378174, "num_tokens": 716974011.0, "step": 18789 }, { "epoch": 2.390281134715685, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.913447380065918, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8733731508255005, "num_tokens": 717012438.0, "step": 18790 }, { "epoch": 2.3904083449942757, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.882214546203613, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8643683195114136, "num_tokens": 717051755.0, "step": 18791 }, { "epoch": 2.390535555272866, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.866190910339355, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.873613715171814, "num_tokens": 717090740.0, "step": 18792 }, { "epoch": 2.3906627655514567, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.935962677001953, "learning_rate": 1e-06, "loss": 0.4626, "mean_token_accuracy": 0.8484549522399902, "num_tokens": 717132033.0, "step": 18793 }, { "epoch": 2.390789975830047, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.85837173461914, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8800955414772034, "num_tokens": 717168255.0, "step": 18794 }, { "epoch": 2.390917186108638, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.937749862670898, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8714569807052612, "num_tokens": 717206863.0, "step": 18795 }, { "epoch": 2.391044396387228, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.89755630493164, "learning_rate": 1e-06, "loss": 0.4608, "mean_token_accuracy": 0.8531087636947632, "num_tokens": 717245374.0, "step": 18796 }, { "epoch": 2.3911716066658184, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.877288818359375, "learning_rate": 1e-06, "loss": 0.3459, "mean_token_accuracy": 0.8911803364753723, "num_tokens": 717279821.0, "step": 18797 }, { "epoch": 2.391298816944409, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.983322143554688, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8661936521530151, "num_tokens": 717318896.0, "step": 18798 }, { "epoch": 2.3914260272229995, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.922593116760254, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8639214038848877, "num_tokens": 717361493.0, "step": 18799 }, { "epoch": 2.39155323750159, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.928631782531738, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8652002811431885, "num_tokens": 717402801.0, "step": 18800 }, { "epoch": 2.3916804477801805, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.904325485229492, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8908973932266235, "num_tokens": 717441275.0, "step": 18801 }, { "epoch": 2.391807658058771, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.897309303283691, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8615018129348755, "num_tokens": 717477423.0, "step": 18802 }, { "epoch": 2.3919348683373616, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.917044639587402, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8665198087692261, "num_tokens": 717512978.0, "step": 18803 }, { "epoch": 2.392062078615952, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.879406929016113, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8918282985687256, "num_tokens": 717549710.0, "step": 18804 }, { "epoch": 2.3921892888945426, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.924710273742676, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8689668774604797, "num_tokens": 717586588.0, "step": 18805 }, { "epoch": 2.392316499173133, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.870265007019043, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.886035144329071, "num_tokens": 717621507.0, "step": 18806 }, { "epoch": 2.3924437094517237, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.887386322021484, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8754100799560547, "num_tokens": 717655915.0, "step": 18807 }, { "epoch": 2.392570919730314, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.918469429016113, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8705659508705139, "num_tokens": 717696357.0, "step": 18808 }, { "epoch": 2.3926981300089047, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.886519432067871, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8863979578018188, "num_tokens": 717735397.0, "step": 18809 }, { "epoch": 2.3928253402874953, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.880169868469238, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.877457857131958, "num_tokens": 717770081.0, "step": 18810 }, { "epoch": 2.392952550566086, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.899680137634277, "learning_rate": 1e-06, "loss": 0.3441, "mean_token_accuracy": 0.8901144862174988, "num_tokens": 717805262.0, "step": 18811 }, { "epoch": 2.3930797608446763, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.988072395324707, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8731421232223511, "num_tokens": 717841877.0, "step": 18812 }, { "epoch": 2.393206971123267, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.867490768432617, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8577172756195068, "num_tokens": 717881435.0, "step": 18813 }, { "epoch": 2.3933341814018574, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.900592803955078, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.892507791519165, "num_tokens": 717922240.0, "step": 18814 }, { "epoch": 2.393461391680448, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.894693374633789, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8702251315116882, "num_tokens": 717958438.0, "step": 18815 }, { "epoch": 2.3935886019590384, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.950156211853027, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8648138046264648, "num_tokens": 717994451.0, "step": 18816 }, { "epoch": 2.393715812237629, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.944086074829102, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8754909634590149, "num_tokens": 718034318.0, "step": 18817 }, { "epoch": 2.3938430225162195, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.899199485778809, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8725048899650574, "num_tokens": 718074994.0, "step": 18818 }, { "epoch": 2.3939702327948096, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.937479019165039, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8705262541770935, "num_tokens": 718118849.0, "step": 18819 }, { "epoch": 2.3940974430734006, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.906452178955078, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8793714046478271, "num_tokens": 718152874.0, "step": 18820 }, { "epoch": 2.3942246533519906, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.897348403930664, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8667157888412476, "num_tokens": 718194206.0, "step": 18821 }, { "epoch": 2.394351863630581, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.874534606933594, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8768346309661865, "num_tokens": 718227408.0, "step": 18822 }, { "epoch": 2.3944790739091717, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.92234992980957, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8677673935890198, "num_tokens": 718266962.0, "step": 18823 }, { "epoch": 2.3946062841877622, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.001753807067871, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8788617253303528, "num_tokens": 718307873.0, "step": 18824 }, { "epoch": 2.3947334944663528, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.895572662353516, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8792562484741211, "num_tokens": 718350296.0, "step": 18825 }, { "epoch": 2.3948607047449433, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.89846134185791, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.876239001750946, "num_tokens": 718390734.0, "step": 18826 }, { "epoch": 2.394987915023534, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.876852035522461, "learning_rate": 1e-06, "loss": 0.3508, "mean_token_accuracy": 0.889484703540802, "num_tokens": 718435630.0, "step": 18827 }, { "epoch": 2.3951151253021243, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.922791481018066, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8808249235153198, "num_tokens": 718474791.0, "step": 18828 }, { "epoch": 2.395242335580715, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.962661743164062, "learning_rate": 1e-06, "loss": 0.4544, "mean_token_accuracy": 0.8566323518753052, "num_tokens": 718514957.0, "step": 18829 }, { "epoch": 2.3953695458593054, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.880412101745605, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8849004507064819, "num_tokens": 718551761.0, "step": 18830 }, { "epoch": 2.395496756137896, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.96835708618164, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8729010820388794, "num_tokens": 718586445.0, "step": 18831 }, { "epoch": 2.3956239664164865, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.92876148223877, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8691011071205139, "num_tokens": 718631105.0, "step": 18832 }, { "epoch": 2.395751176695077, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.94040298461914, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8699907064437866, "num_tokens": 718672278.0, "step": 18833 }, { "epoch": 2.3958783869736675, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.917267799377441, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.879348635673523, "num_tokens": 718713191.0, "step": 18834 }, { "epoch": 2.396005597252258, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.90248966217041, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8669630289077759, "num_tokens": 718745012.0, "step": 18835 }, { "epoch": 2.3961328075308486, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.949942588806152, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.873887836933136, "num_tokens": 718780149.0, "step": 18836 }, { "epoch": 2.396260017809439, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.920652389526367, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8693402409553528, "num_tokens": 718812831.0, "step": 18837 }, { "epoch": 2.3963872280880296, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.914306640625, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.884239137172699, "num_tokens": 718847650.0, "step": 18838 }, { "epoch": 2.39651443836662, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.931784629821777, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8751698136329651, "num_tokens": 718883417.0, "step": 18839 }, { "epoch": 2.3966416486452107, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.89873218536377, "learning_rate": 1e-06, "loss": 0.3176, "mean_token_accuracy": 0.8982312679290771, "num_tokens": 718915501.0, "step": 18840 }, { "epoch": 2.396768858923801, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.927347183227539, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8873358964920044, "num_tokens": 718954485.0, "step": 18841 }, { "epoch": 2.3968960692023917, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.924063682556152, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8589491844177246, "num_tokens": 718995815.0, "step": 18842 }, { "epoch": 2.3970232794809823, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.915762901306152, "learning_rate": 1e-06, "loss": 0.3364, "mean_token_accuracy": 0.8880149126052856, "num_tokens": 719032820.0, "step": 18843 }, { "epoch": 2.3971504897595723, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 14.000798225402832, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.871675968170166, "num_tokens": 719066966.0, "step": 18844 }, { "epoch": 2.3972777000381633, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.990185737609863, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.873582124710083, "num_tokens": 719108768.0, "step": 18845 }, { "epoch": 2.3974049103167534, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.925846099853516, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8712153434753418, "num_tokens": 719151451.0, "step": 18846 }, { "epoch": 2.397532120595344, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.974810600280762, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8655381202697754, "num_tokens": 719192103.0, "step": 18847 }, { "epoch": 2.3976593308739345, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.957880020141602, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8701996207237244, "num_tokens": 719226514.0, "step": 18848 }, { "epoch": 2.397786541152525, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.992913246154785, "learning_rate": 1e-06, "loss": 0.3243, "mean_token_accuracy": 0.8969087600708008, "num_tokens": 719262944.0, "step": 18849 }, { "epoch": 2.3979137514311155, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.931097984313965, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8801687359809875, "num_tokens": 719306707.0, "step": 18850 }, { "epoch": 2.398040961709706, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.895371437072754, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8791579008102417, "num_tokens": 719347523.0, "step": 18851 }, { "epoch": 2.3981681719882966, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.986788749694824, "learning_rate": 1e-06, "loss": 0.3296, "mean_token_accuracy": 0.8932698965072632, "num_tokens": 719379474.0, "step": 18852 }, { "epoch": 2.398295382266887, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.933269500732422, "learning_rate": 1e-06, "loss": 0.4527, "mean_token_accuracy": 0.8583003282546997, "num_tokens": 719421624.0, "step": 18853 }, { "epoch": 2.3984225925454776, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.966897010803223, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8693940043449402, "num_tokens": 719455570.0, "step": 18854 }, { "epoch": 2.398549802824068, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.916396141052246, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8908511400222778, "num_tokens": 719500300.0, "step": 18855 }, { "epoch": 2.3986770131026587, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.965933799743652, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.858790397644043, "num_tokens": 719538119.0, "step": 18856 }, { "epoch": 2.398804223381249, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.904388427734375, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8654404878616333, "num_tokens": 719574012.0, "step": 18857 }, { "epoch": 2.3989314336598397, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.99848461151123, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.885125458240509, "num_tokens": 719609548.0, "step": 18858 }, { "epoch": 2.3990586439384303, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.944422721862793, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8677242398262024, "num_tokens": 719647280.0, "step": 18859 }, { "epoch": 2.399185854217021, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.886052131652832, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.8885189294815063, "num_tokens": 719686987.0, "step": 18860 }, { "epoch": 2.3993130644956113, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.99988842010498, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.8925591707229614, "num_tokens": 719727974.0, "step": 18861 }, { "epoch": 2.399440274774202, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.959186553955078, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8744828104972839, "num_tokens": 719764308.0, "step": 18862 }, { "epoch": 2.3995674850527924, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.974808692932129, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8636070489883423, "num_tokens": 719801872.0, "step": 18863 }, { "epoch": 2.399694695331383, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.912028312683105, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.876673698425293, "num_tokens": 719842504.0, "step": 18864 }, { "epoch": 2.3998219056099734, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.941431999206543, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8686085939407349, "num_tokens": 719878885.0, "step": 18865 }, { "epoch": 2.399949115888564, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.878521919250488, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8581398129463196, "num_tokens": 719922891.0, "step": 18866 }, { "epoch": 2.4000763261671545, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.914801597595215, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8722891807556152, "num_tokens": 719960113.0, "step": 18867 }, { "epoch": 2.400203536445745, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.889335632324219, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8731351494789124, "num_tokens": 720006079.0, "step": 18868 }, { "epoch": 2.400330746724335, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.915191650390625, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8615767955780029, "num_tokens": 720042219.0, "step": 18869 }, { "epoch": 2.400457957002926, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.871954917907715, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.874462366104126, "num_tokens": 720082574.0, "step": 18870 }, { "epoch": 2.400585167281516, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.972455024719238, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8728047609329224, "num_tokens": 720115583.0, "step": 18871 }, { "epoch": 2.4007123775601067, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.982123374938965, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8655652403831482, "num_tokens": 720151350.0, "step": 18872 }, { "epoch": 2.400839587838697, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.906805038452148, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8665956258773804, "num_tokens": 720187072.0, "step": 18873 }, { "epoch": 2.4009667981172877, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.977005958557129, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8774895668029785, "num_tokens": 720222249.0, "step": 18874 }, { "epoch": 2.4010940083958783, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.919312477111816, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8725259304046631, "num_tokens": 720264682.0, "step": 18875 }, { "epoch": 2.401221218674469, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.914742469787598, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8705254793167114, "num_tokens": 720309935.0, "step": 18876 }, { "epoch": 2.4013484289530593, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.914413452148438, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.869610607624054, "num_tokens": 720343992.0, "step": 18877 }, { "epoch": 2.40147563923165, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.93786334991455, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8728094100952148, "num_tokens": 720379707.0, "step": 18878 }, { "epoch": 2.4016028495102404, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.90684700012207, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8706622123718262, "num_tokens": 720419796.0, "step": 18879 }, { "epoch": 2.401730059788831, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.940897941589355, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8762087225914001, "num_tokens": 720460919.0, "step": 18880 }, { "epoch": 2.4018572700674214, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.917572021484375, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8701000213623047, "num_tokens": 720498403.0, "step": 18881 }, { "epoch": 2.401984480346012, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.853952407836914, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8743084669113159, "num_tokens": 720534931.0, "step": 18882 }, { "epoch": 2.4021116906246025, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.951090812683105, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8768346309661865, "num_tokens": 720574062.0, "step": 18883 }, { "epoch": 2.402238900903193, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.958097457885742, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8652927875518799, "num_tokens": 720611552.0, "step": 18884 }, { "epoch": 2.4023661111817836, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.899158477783203, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8733865022659302, "num_tokens": 720647886.0, "step": 18885 }, { "epoch": 2.402493321460374, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.918681144714355, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8791625499725342, "num_tokens": 720688034.0, "step": 18886 }, { "epoch": 2.4026205317389646, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.899853706359863, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8799192905426025, "num_tokens": 720727805.0, "step": 18887 }, { "epoch": 2.402747742017555, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.965927124023438, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8763407468795776, "num_tokens": 720759717.0, "step": 18888 }, { "epoch": 2.4028749522961457, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.936653137207031, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.867448627948761, "num_tokens": 720790435.0, "step": 18889 }, { "epoch": 2.403002162574736, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.977983474731445, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8695358633995056, "num_tokens": 720831715.0, "step": 18890 }, { "epoch": 2.4031293728533267, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.932526588439941, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8725935220718384, "num_tokens": 720869598.0, "step": 18891 }, { "epoch": 2.403256583131917, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.994649887084961, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8682435750961304, "num_tokens": 720911122.0, "step": 18892 }, { "epoch": 2.403383793410508, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.94393253326416, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8771514892578125, "num_tokens": 720952265.0, "step": 18893 }, { "epoch": 2.403511003689098, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.938802719116211, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8778545260429382, "num_tokens": 720991191.0, "step": 18894 }, { "epoch": 2.4036382139676884, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.000873565673828, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8836387395858765, "num_tokens": 721032524.0, "step": 18895 }, { "epoch": 2.403765424246279, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.948506355285645, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8725343346595764, "num_tokens": 721072699.0, "step": 18896 }, { "epoch": 2.4038926345248695, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.909255981445312, "learning_rate": 1e-06, "loss": 0.3545, "mean_token_accuracy": 0.8850445747375488, "num_tokens": 721109467.0, "step": 18897 }, { "epoch": 2.40401984480346, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.978836059570312, "learning_rate": 1e-06, "loss": 0.4827, "mean_token_accuracy": 0.8501445651054382, "num_tokens": 721142108.0, "step": 18898 }, { "epoch": 2.4041470550820505, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.943160057067871, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8666841983795166, "num_tokens": 721179875.0, "step": 18899 }, { "epoch": 2.404274265360641, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.920060157775879, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8710006475448608, "num_tokens": 721217141.0, "step": 18900 }, { "epoch": 2.4044014756392316, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.932560920715332, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8673208951950073, "num_tokens": 721256357.0, "step": 18901 }, { "epoch": 2.404528685917822, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.973673820495605, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8726928234100342, "num_tokens": 721292610.0, "step": 18902 }, { "epoch": 2.4046558961964126, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.972884178161621, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8795897960662842, "num_tokens": 721329048.0, "step": 18903 }, { "epoch": 2.404783106475003, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.967680931091309, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.8886171579360962, "num_tokens": 721364249.0, "step": 18904 }, { "epoch": 2.4049103167535937, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.032567024230957, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8806816339492798, "num_tokens": 721391154.0, "step": 18905 }, { "epoch": 2.405037527032184, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.987605094909668, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8839487433433533, "num_tokens": 721428443.0, "step": 18906 }, { "epoch": 2.4051647373107747, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.964223861694336, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.863552451133728, "num_tokens": 721474545.0, "step": 18907 }, { "epoch": 2.4052919475893653, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.936959266662598, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8743804693222046, "num_tokens": 721513153.0, "step": 18908 }, { "epoch": 2.405419157867956, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.932971954345703, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.870388388633728, "num_tokens": 721556287.0, "step": 18909 }, { "epoch": 2.4055463681465463, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.9873628616333, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8800891637802124, "num_tokens": 721590219.0, "step": 18910 }, { "epoch": 2.405673578425137, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.945808410644531, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8730164766311646, "num_tokens": 721627308.0, "step": 18911 }, { "epoch": 2.4058007887037274, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.993505477905273, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8690446615219116, "num_tokens": 721664218.0, "step": 18912 }, { "epoch": 2.405927998982318, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.914212226867676, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8846402168273926, "num_tokens": 721703511.0, "step": 18913 }, { "epoch": 2.4060552092609084, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.003299713134766, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8861163854598999, "num_tokens": 721741512.0, "step": 18914 }, { "epoch": 2.406182419539499, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.942586898803711, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8775559663772583, "num_tokens": 721778443.0, "step": 18915 }, { "epoch": 2.4063096298180895, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.984184265136719, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8703290820121765, "num_tokens": 721817226.0, "step": 18916 }, { "epoch": 2.4064368400966796, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.975961685180664, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8600008487701416, "num_tokens": 721859419.0, "step": 18917 }, { "epoch": 2.4065640503752705, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.890756607055664, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8871259689331055, "num_tokens": 721895183.0, "step": 18918 }, { "epoch": 2.4066912606538606, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.993038177490234, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8597989082336426, "num_tokens": 721933566.0, "step": 18919 }, { "epoch": 2.406818470932451, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.896162986755371, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8777399659156799, "num_tokens": 721968827.0, "step": 18920 }, { "epoch": 2.4069456812110417, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.924066543579102, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8769668340682983, "num_tokens": 722004989.0, "step": 18921 }, { "epoch": 2.407072891489632, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.9374418258667, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8763530254364014, "num_tokens": 722043681.0, "step": 18922 }, { "epoch": 2.4072001017682227, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.94726848602295, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8740540742874146, "num_tokens": 722086541.0, "step": 18923 }, { "epoch": 2.4073273120468133, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.949100494384766, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8744818568229675, "num_tokens": 722125519.0, "step": 18924 }, { "epoch": 2.407454522325404, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.914800643920898, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8696478009223938, "num_tokens": 722157665.0, "step": 18925 }, { "epoch": 2.4075817326039943, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.996248245239258, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8861660957336426, "num_tokens": 722191819.0, "step": 18926 }, { "epoch": 2.407708942882585, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.947929382324219, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8629828691482544, "num_tokens": 722229755.0, "step": 18927 }, { "epoch": 2.4078361531611754, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.01209545135498, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8634025454521179, "num_tokens": 722267440.0, "step": 18928 }, { "epoch": 2.407963363439766, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.881954193115234, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8692284226417542, "num_tokens": 722310102.0, "step": 18929 }, { "epoch": 2.4080905737183564, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.997264862060547, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8674750328063965, "num_tokens": 722346713.0, "step": 18930 }, { "epoch": 2.408217783996947, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.932947158813477, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8687538504600525, "num_tokens": 722387803.0, "step": 18931 }, { "epoch": 2.4083449942755375, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.916109085083008, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8647088408470154, "num_tokens": 722428943.0, "step": 18932 }, { "epoch": 2.408472204554128, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.976064682006836, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8674935102462769, "num_tokens": 722467860.0, "step": 18933 }, { "epoch": 2.4085994148327186, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.942855834960938, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8689197897911072, "num_tokens": 722503067.0, "step": 18934 }, { "epoch": 2.408726625111309, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.957151412963867, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8759931325912476, "num_tokens": 722537842.0, "step": 18935 }, { "epoch": 2.4088538353898996, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.936018943786621, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8798392415046692, "num_tokens": 722574036.0, "step": 18936 }, { "epoch": 2.40898104566849, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.980545043945312, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8743351697921753, "num_tokens": 722606087.0, "step": 18937 }, { "epoch": 2.4091082559470807, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.922316551208496, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8721647262573242, "num_tokens": 722648745.0, "step": 18938 }, { "epoch": 2.409235466225671, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.857831954956055, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8656936287879944, "num_tokens": 722688589.0, "step": 18939 }, { "epoch": 2.4093626765042617, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.96085262298584, "learning_rate": 1e-06, "loss": 0.3496, "mean_token_accuracy": 0.8905918598175049, "num_tokens": 722723941.0, "step": 18940 }, { "epoch": 2.4094898867828523, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.868706703186035, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8564455509185791, "num_tokens": 722768242.0, "step": 18941 }, { "epoch": 2.4096170970614423, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.977259635925293, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8734363317489624, "num_tokens": 722800970.0, "step": 18942 }, { "epoch": 2.4097443073400333, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.988475799560547, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8649964928627014, "num_tokens": 722839168.0, "step": 18943 }, { "epoch": 2.4098715176186234, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.894129753112793, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8844361901283264, "num_tokens": 722883451.0, "step": 18944 }, { "epoch": 2.409998727897214, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.931654930114746, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8746387362480164, "num_tokens": 722928581.0, "step": 18945 }, { "epoch": 2.4101259381758044, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.938404083251953, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.884777843952179, "num_tokens": 722967141.0, "step": 18946 }, { "epoch": 2.410253148454395, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.001009941101074, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8706189393997192, "num_tokens": 723006263.0, "step": 18947 }, { "epoch": 2.4103803587329855, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.95905876159668, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8767224550247192, "num_tokens": 723050843.0, "step": 18948 }, { "epoch": 2.410507569011576, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.93985366821289, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8741953372955322, "num_tokens": 723092124.0, "step": 18949 }, { "epoch": 2.4106347792901666, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.026976585388184, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8845536708831787, "num_tokens": 723122478.0, "step": 18950 }, { "epoch": 2.410761989568757, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.397628784179688, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8682403564453125, "num_tokens": 723163054.0, "step": 18951 }, { "epoch": 2.4108891998473476, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.97785758972168, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8734732270240784, "num_tokens": 723201518.0, "step": 18952 }, { "epoch": 2.411016410125938, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 13.909830093383789, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.873624324798584, "num_tokens": 723235685.0, "step": 18953 }, { "epoch": 2.4111436204045287, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 14.055742263793945, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.873940110206604, "num_tokens": 723266510.0, "step": 18954 }, { "epoch": 2.411270830683119, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.897676467895508, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8767887353897095, "num_tokens": 723302800.0, "step": 18955 }, { "epoch": 2.4113980409617097, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.949723243713379, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8762617111206055, "num_tokens": 723340669.0, "step": 18956 }, { "epoch": 2.4115252512403003, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.932987213134766, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.877428412437439, "num_tokens": 723377556.0, "step": 18957 }, { "epoch": 2.411652461518891, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.05383586883545, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.878650426864624, "num_tokens": 723410942.0, "step": 18958 }, { "epoch": 2.4117796717974813, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.93203353881836, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.886512041091919, "num_tokens": 723451131.0, "step": 18959 }, { "epoch": 2.411906882076072, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.967394828796387, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8775687217712402, "num_tokens": 723485876.0, "step": 18960 }, { "epoch": 2.4120340923546624, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.950017929077148, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8820391893386841, "num_tokens": 723521252.0, "step": 18961 }, { "epoch": 2.412161302633253, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.952095031738281, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8585779666900635, "num_tokens": 723561094.0, "step": 18962 }, { "epoch": 2.4122885129118434, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.983728408813477, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8728508949279785, "num_tokens": 723597281.0, "step": 18963 }, { "epoch": 2.412415723190434, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.971821784973145, "learning_rate": 1e-06, "loss": 0.3261, "mean_token_accuracy": 0.8961136341094971, "num_tokens": 723637690.0, "step": 18964 }, { "epoch": 2.4125429334690245, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.960901260375977, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8740507364273071, "num_tokens": 723676250.0, "step": 18965 }, { "epoch": 2.412670143747615, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.03097152709961, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8836917877197266, "num_tokens": 723713157.0, "step": 18966 }, { "epoch": 2.412797354026205, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.985255241394043, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8789190053939819, "num_tokens": 723745877.0, "step": 18967 }, { "epoch": 2.412924564304796, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.98951244354248, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8758392930030823, "num_tokens": 723784285.0, "step": 18968 }, { "epoch": 2.413051774583386, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.973517417907715, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8631327748298645, "num_tokens": 723817085.0, "step": 18969 }, { "epoch": 2.4131789848619767, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.93045711517334, "learning_rate": 1e-06, "loss": 0.4226, "mean_token_accuracy": 0.8697113990783691, "num_tokens": 723858466.0, "step": 18970 }, { "epoch": 2.413306195140567, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.965108871459961, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8731364011764526, "num_tokens": 723896627.0, "step": 18971 }, { "epoch": 2.4134334054191577, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.941704750061035, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8688254952430725, "num_tokens": 723930727.0, "step": 18972 }, { "epoch": 2.4135606156977483, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.008259773254395, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8637338876724243, "num_tokens": 723965996.0, "step": 18973 }, { "epoch": 2.413687825976339, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.979899406433105, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8688316941261292, "num_tokens": 724002441.0, "step": 18974 }, { "epoch": 2.4138150362549293, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.99094295501709, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8777872323989868, "num_tokens": 724039238.0, "step": 18975 }, { "epoch": 2.41394224653352, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.976729393005371, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8791245222091675, "num_tokens": 724077430.0, "step": 18976 }, { "epoch": 2.4140694568121104, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.00331974029541, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8701256513595581, "num_tokens": 724113175.0, "step": 18977 }, { "epoch": 2.414196667090701, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.945042610168457, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8685649037361145, "num_tokens": 724156409.0, "step": 18978 }, { "epoch": 2.4143238773692914, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.97196102142334, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8747135996818542, "num_tokens": 724199879.0, "step": 18979 }, { "epoch": 2.414451087647882, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.004683494567871, "learning_rate": 1e-06, "loss": 0.3525, "mean_token_accuracy": 0.8869807720184326, "num_tokens": 724238065.0, "step": 18980 }, { "epoch": 2.4145782979264725, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.943487167358398, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8668078184127808, "num_tokens": 724272336.0, "step": 18981 }, { "epoch": 2.414705508205063, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.071757316589355, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8658021688461304, "num_tokens": 724305200.0, "step": 18982 }, { "epoch": 2.4148327184836536, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.981793403625488, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8784087300300598, "num_tokens": 724341295.0, "step": 18983 }, { "epoch": 2.414959928762244, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.949917793273926, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8764567971229553, "num_tokens": 724384336.0, "step": 18984 }, { "epoch": 2.4150871390408346, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.051087379455566, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8704513311386108, "num_tokens": 724417279.0, "step": 18985 }, { "epoch": 2.415214349319425, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.96995735168457, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.872873842716217, "num_tokens": 724456509.0, "step": 18986 }, { "epoch": 2.4153415595980157, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.981522560119629, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8839117288589478, "num_tokens": 724498812.0, "step": 18987 }, { "epoch": 2.415468769876606, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.948742866516113, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.8613041639328003, "num_tokens": 724538174.0, "step": 18988 }, { "epoch": 2.4155959801551967, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.92795181274414, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8791729211807251, "num_tokens": 724577171.0, "step": 18989 }, { "epoch": 2.415723190433787, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.97115707397461, "learning_rate": 1e-06, "loss": 0.3333, "mean_token_accuracy": 0.8944625854492188, "num_tokens": 724610626.0, "step": 18990 }, { "epoch": 2.4158504007123778, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.925210952758789, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.885589063167572, "num_tokens": 724647613.0, "step": 18991 }, { "epoch": 2.415977610990968, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.937274932861328, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8757616281509399, "num_tokens": 724684898.0, "step": 18992 }, { "epoch": 2.4161048212695584, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.912508964538574, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8698458075523376, "num_tokens": 724725566.0, "step": 18993 }, { "epoch": 2.416232031548149, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.926578521728516, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8647480010986328, "num_tokens": 724763701.0, "step": 18994 }, { "epoch": 2.4163592418267394, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.934342384338379, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8733193874359131, "num_tokens": 724798108.0, "step": 18995 }, { "epoch": 2.41648645210533, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.929086685180664, "learning_rate": 1e-06, "loss": 0.4537, "mean_token_accuracy": 0.853495717048645, "num_tokens": 724842430.0, "step": 18996 }, { "epoch": 2.4166136623839205, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.950961112976074, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8697373867034912, "num_tokens": 724884585.0, "step": 18997 }, { "epoch": 2.416740872662511, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.943462371826172, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8850171566009521, "num_tokens": 724919338.0, "step": 18998 }, { "epoch": 2.4168680829411016, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.97464656829834, "learning_rate": 1e-06, "loss": 0.3369, "mean_token_accuracy": 0.8950296640396118, "num_tokens": 724956007.0, "step": 18999 }, { "epoch": 2.416995293219692, "ewc_loss": 0.03662109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 13.945771217346191, "learning_rate": 1e-06, "loss": 0.3486, "mean_token_accuracy": 0.8904088735580444, "num_tokens": 724996994.0, "step": 19000 }, { "epoch": 2.4171225034982826, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.920222282409668, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8680453300476074, "num_tokens": 725035607.0, "step": 19001 }, { "epoch": 2.417249713776873, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.99199104309082, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8776005506515503, "num_tokens": 725075370.0, "step": 19002 }, { "epoch": 2.4173769240554637, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.899767875671387, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8716111779212952, "num_tokens": 725119536.0, "step": 19003 }, { "epoch": 2.417504134334054, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.953782081604004, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8737040758132935, "num_tokens": 725158048.0, "step": 19004 }, { "epoch": 2.4176313446126447, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.930731773376465, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8828110694885254, "num_tokens": 725196343.0, "step": 19005 }, { "epoch": 2.4177585548912353, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.93132495880127, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8747480511665344, "num_tokens": 725234613.0, "step": 19006 }, { "epoch": 2.417885765169826, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.921418190002441, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8728610873222351, "num_tokens": 725272129.0, "step": 19007 }, { "epoch": 2.4180129754484163, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.019923210144043, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8652817010879517, "num_tokens": 725305740.0, "step": 19008 }, { "epoch": 2.418140185727007, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.91899299621582, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8704531192779541, "num_tokens": 725340579.0, "step": 19009 }, { "epoch": 2.4182673960055974, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.986361503601074, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8771312236785889, "num_tokens": 725380492.0, "step": 19010 }, { "epoch": 2.418394606284188, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.956207275390625, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8762882351875305, "num_tokens": 725418043.0, "step": 19011 }, { "epoch": 2.4185218165627784, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.043477058410645, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8659836053848267, "num_tokens": 725456260.0, "step": 19012 }, { "epoch": 2.418649026841369, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.981762886047363, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.870781421661377, "num_tokens": 725496605.0, "step": 19013 }, { "epoch": 2.4187762371199595, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.973260879516602, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8755959868431091, "num_tokens": 725532395.0, "step": 19014 }, { "epoch": 2.4189034473985496, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.937344551086426, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.868201494216919, "num_tokens": 725575335.0, "step": 19015 }, { "epoch": 2.4190306576771405, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.95661449432373, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8802487850189209, "num_tokens": 725612345.0, "step": 19016 }, { "epoch": 2.4191578679557306, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.968169212341309, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8775783777236938, "num_tokens": 725645566.0, "step": 19017 }, { "epoch": 2.419285078234321, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.959809303283691, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8687007427215576, "num_tokens": 725684218.0, "step": 19018 }, { "epoch": 2.4194122885129117, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.972801208496094, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8779879808425903, "num_tokens": 725718478.0, "step": 19019 }, { "epoch": 2.419539498791502, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.95030403137207, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8680028319358826, "num_tokens": 725756208.0, "step": 19020 }, { "epoch": 2.4196667090700927, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.047049522399902, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8677903413772583, "num_tokens": 725791884.0, "step": 19021 }, { "epoch": 2.4197939193486833, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.938094139099121, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8780845403671265, "num_tokens": 725829138.0, "step": 19022 }, { "epoch": 2.419921129627274, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.893595695495605, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8748930096626282, "num_tokens": 725868301.0, "step": 19023 }, { "epoch": 2.4200483399058643, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.959386825561523, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8633707761764526, "num_tokens": 725906186.0, "step": 19024 }, { "epoch": 2.420175550184455, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.936881065368652, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8621991872787476, "num_tokens": 725947006.0, "step": 19025 }, { "epoch": 2.4203027604630454, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.923551559448242, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.878018319606781, "num_tokens": 725987363.0, "step": 19026 }, { "epoch": 2.420429970741636, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.987180709838867, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8646386861801147, "num_tokens": 726026176.0, "step": 19027 }, { "epoch": 2.4205571810202264, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.886011123657227, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8726262450218201, "num_tokens": 726068124.0, "step": 19028 }, { "epoch": 2.420684391298817, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.993398666381836, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8648035526275635, "num_tokens": 726110270.0, "step": 19029 }, { "epoch": 2.4208116015774075, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.99527359008789, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8889976739883423, "num_tokens": 726143908.0, "step": 19030 }, { "epoch": 2.420938811855998, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.985333442687988, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8721081018447876, "num_tokens": 726179082.0, "step": 19031 }, { "epoch": 2.4210660221345885, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.991271018981934, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8666693568229675, "num_tokens": 726216387.0, "step": 19032 }, { "epoch": 2.421193232413179, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.90141773223877, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8829351663589478, "num_tokens": 726258831.0, "step": 19033 }, { "epoch": 2.4213204426917696, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.98056697845459, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8759121894836426, "num_tokens": 726301929.0, "step": 19034 }, { "epoch": 2.42144765297036, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.966668128967285, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8771488666534424, "num_tokens": 726339938.0, "step": 19035 }, { "epoch": 2.4215748632489507, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.966221809387207, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8749717473983765, "num_tokens": 726377197.0, "step": 19036 }, { "epoch": 2.421702073527541, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.932134628295898, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8774126172065735, "num_tokens": 726417591.0, "step": 19037 }, { "epoch": 2.4218292838061317, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.986549377441406, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8636929392814636, "num_tokens": 726456094.0, "step": 19038 }, { "epoch": 2.4219564940847222, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.91999340057373, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8827941417694092, "num_tokens": 726491158.0, "step": 19039 }, { "epoch": 2.4220837043633123, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.948976516723633, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8595764636993408, "num_tokens": 726531334.0, "step": 19040 }, { "epoch": 2.4222109146419033, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.941499710083008, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.864690899848938, "num_tokens": 726571999.0, "step": 19041 }, { "epoch": 2.4223381249204934, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.951900482177734, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8652572631835938, "num_tokens": 726604945.0, "step": 19042 }, { "epoch": 2.422465335199084, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.972550392150879, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8653758764266968, "num_tokens": 726645023.0, "step": 19043 }, { "epoch": 2.4225925454776744, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.922019004821777, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8728429675102234, "num_tokens": 726684297.0, "step": 19044 }, { "epoch": 2.422719755756265, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.88509750366211, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8629637360572815, "num_tokens": 726726950.0, "step": 19045 }, { "epoch": 2.4228469660348555, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.006143569946289, "learning_rate": 1e-06, "loss": 0.4646, "mean_token_accuracy": 0.8552159070968628, "num_tokens": 726761944.0, "step": 19046 }, { "epoch": 2.422974176313446, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.886382102966309, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.883729100227356, "num_tokens": 726799041.0, "step": 19047 }, { "epoch": 2.4231013865920366, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.977163314819336, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8676342964172363, "num_tokens": 726838160.0, "step": 19048 }, { "epoch": 2.423228596870627, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.918188095092773, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8832278251647949, "num_tokens": 726873150.0, "step": 19049 }, { "epoch": 2.4233558071492176, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.952900886535645, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8775445222854614, "num_tokens": 726905253.0, "step": 19050 }, { "epoch": 2.423483017427808, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.998600006103516, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.8840875625610352, "num_tokens": 726942842.0, "step": 19051 }, { "epoch": 2.4236102277063987, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.916379928588867, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.878158450126648, "num_tokens": 726987974.0, "step": 19052 }, { "epoch": 2.423737437984989, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.943257331848145, "learning_rate": 1e-06, "loss": 0.3449, "mean_token_accuracy": 0.8916236758232117, "num_tokens": 727021614.0, "step": 19053 }, { "epoch": 2.4238646482635797, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.96791934967041, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8726556301116943, "num_tokens": 727061773.0, "step": 19054 }, { "epoch": 2.4239918585421703, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.003273963928223, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8687450289726257, "num_tokens": 727095877.0, "step": 19055 }, { "epoch": 2.424119068820761, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.998538970947266, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8763915300369263, "num_tokens": 727128676.0, "step": 19056 }, { "epoch": 2.4242462790993513, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.94107437133789, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8833919167518616, "num_tokens": 727164528.0, "step": 19057 }, { "epoch": 2.424373489377942, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.905794143676758, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8803951740264893, "num_tokens": 727202403.0, "step": 19058 }, { "epoch": 2.4245006996565324, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.957398414611816, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.876548707485199, "num_tokens": 727242231.0, "step": 19059 }, { "epoch": 2.424627909935123, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.925789833068848, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8612119555473328, "num_tokens": 727276371.0, "step": 19060 }, { "epoch": 2.4247551202137134, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.941034317016602, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.877666711807251, "num_tokens": 727320101.0, "step": 19061 }, { "epoch": 2.424882330492304, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.970498085021973, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8654465079307556, "num_tokens": 727363687.0, "step": 19062 }, { "epoch": 2.4250095407708945, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.958107948303223, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8797710537910461, "num_tokens": 727399772.0, "step": 19063 }, { "epoch": 2.425136751049485, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.964412689208984, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8819543719291687, "num_tokens": 727442944.0, "step": 19064 }, { "epoch": 2.425263961328075, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.009385108947754, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.881989598274231, "num_tokens": 727483283.0, "step": 19065 }, { "epoch": 2.425391171606666, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.977092742919922, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.871599555015564, "num_tokens": 727526316.0, "step": 19066 }, { "epoch": 2.425518381885256, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.937679290771484, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8815068006515503, "num_tokens": 727563218.0, "step": 19067 }, { "epoch": 2.4256455921638467, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.007826805114746, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8860911130905151, "num_tokens": 727594877.0, "step": 19068 }, { "epoch": 2.425772802442437, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.959334373474121, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8870356678962708, "num_tokens": 727627561.0, "step": 19069 }, { "epoch": 2.4259000127210277, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.998100280761719, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8653004169464111, "num_tokens": 727664384.0, "step": 19070 }, { "epoch": 2.4260272229996183, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.97493839263916, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8715792894363403, "num_tokens": 727700866.0, "step": 19071 }, { "epoch": 2.426154433278209, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.975430488586426, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8909738063812256, "num_tokens": 727741539.0, "step": 19072 }, { "epoch": 2.4262816435567993, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.90716552734375, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8748201131820679, "num_tokens": 727777828.0, "step": 19073 }, { "epoch": 2.42640885383539, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.940299987792969, "learning_rate": 1e-06, "loss": 0.3656, "mean_token_accuracy": 0.8823357820510864, "num_tokens": 727818979.0, "step": 19074 }, { "epoch": 2.4265360641139804, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.96861743927002, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8746846318244934, "num_tokens": 727856440.0, "step": 19075 }, { "epoch": 2.426663274392571, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.990813255310059, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8822542428970337, "num_tokens": 727895036.0, "step": 19076 }, { "epoch": 2.4267904846711614, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.975253105163574, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8642945289611816, "num_tokens": 727932292.0, "step": 19077 }, { "epoch": 2.426917694949752, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.94569206237793, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8735637068748474, "num_tokens": 727967005.0, "step": 19078 }, { "epoch": 2.4270449052283425, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.010810852050781, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8651261925697327, "num_tokens": 728004262.0, "step": 19079 }, { "epoch": 2.427172115506933, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.974834442138672, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8836888074874878, "num_tokens": 728041451.0, "step": 19080 }, { "epoch": 2.4272993257855235, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.9351167678833, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8724817037582397, "num_tokens": 728079217.0, "step": 19081 }, { "epoch": 2.427426536064114, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.928372383117676, "learning_rate": 1e-06, "loss": 0.3367, "mean_token_accuracy": 0.8897455334663391, "num_tokens": 728116221.0, "step": 19082 }, { "epoch": 2.4275537463427046, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.967301368713379, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.866273820400238, "num_tokens": 728154861.0, "step": 19083 }, { "epoch": 2.427680956621295, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.973496437072754, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8637209534645081, "num_tokens": 728188276.0, "step": 19084 }, { "epoch": 2.4278081668998857, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.977272033691406, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8851945400238037, "num_tokens": 728229313.0, "step": 19085 }, { "epoch": 2.427935377178476, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.957682609558105, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8636213541030884, "num_tokens": 728268706.0, "step": 19086 }, { "epoch": 2.4280625874570667, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.017722129821777, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8872576355934143, "num_tokens": 728297490.0, "step": 19087 }, { "epoch": 2.428189797735657, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.999571800231934, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.867611289024353, "num_tokens": 728336041.0, "step": 19088 }, { "epoch": 2.4283170080142478, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.93795394897461, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8811066150665283, "num_tokens": 728372701.0, "step": 19089 }, { "epoch": 2.428444218292838, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.988563537597656, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8775511384010315, "num_tokens": 728408138.0, "step": 19090 }, { "epoch": 2.4285714285714284, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.973044395446777, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8808678388595581, "num_tokens": 728452649.0, "step": 19091 }, { "epoch": 2.428698638850019, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.032258033752441, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8640841245651245, "num_tokens": 728490551.0, "step": 19092 }, { "epoch": 2.4288258491286094, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.974434852600098, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8798834085464478, "num_tokens": 728529558.0, "step": 19093 }, { "epoch": 2.4289530594072, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.910239219665527, "learning_rate": 1e-06, "loss": 0.4712, "mean_token_accuracy": 0.8510835766792297, "num_tokens": 728573495.0, "step": 19094 }, { "epoch": 2.4290802696857905, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.976597785949707, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8753103017807007, "num_tokens": 728609166.0, "step": 19095 }, { "epoch": 2.429207479964381, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.986825942993164, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8805742263793945, "num_tokens": 728653070.0, "step": 19096 }, { "epoch": 2.4293346902429716, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.925926208496094, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8772683143615723, "num_tokens": 728689142.0, "step": 19097 }, { "epoch": 2.429461900521562, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.964981079101562, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8784049153327942, "num_tokens": 728725615.0, "step": 19098 }, { "epoch": 2.4295891108001526, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.9994535446167, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.865020215511322, "num_tokens": 728762366.0, "step": 19099 }, { "epoch": 2.429716321078743, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.957897186279297, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8620826005935669, "num_tokens": 728801951.0, "step": 19100 }, { "epoch": 2.4298435313573337, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.965675354003906, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8831820487976074, "num_tokens": 728836403.0, "step": 19101 }, { "epoch": 2.429970741635924, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.94206428527832, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8806664943695068, "num_tokens": 728878485.0, "step": 19102 }, { "epoch": 2.4300979519145147, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.925336837768555, "learning_rate": 1e-06, "loss": 0.3594, "mean_token_accuracy": 0.8827732801437378, "num_tokens": 728915935.0, "step": 19103 }, { "epoch": 2.4302251621931052, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.99240493774414, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.858295202255249, "num_tokens": 728956847.0, "step": 19104 }, { "epoch": 2.4303523724716958, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.962807655334473, "learning_rate": 1e-06, "loss": 0.385, "mean_token_accuracy": 0.8790916204452515, "num_tokens": 728994865.0, "step": 19105 }, { "epoch": 2.4304795827502863, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.99743366241455, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8680170774459839, "num_tokens": 729033030.0, "step": 19106 }, { "epoch": 2.430606793028877, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.944404602050781, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8849467039108276, "num_tokens": 729070430.0, "step": 19107 }, { "epoch": 2.4307340033074674, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.003301620483398, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8856850862503052, "num_tokens": 729102127.0, "step": 19108 }, { "epoch": 2.430861213586058, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.986477851867676, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8775761127471924, "num_tokens": 729138128.0, "step": 19109 }, { "epoch": 2.4309884238646484, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.952799797058105, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8718430995941162, "num_tokens": 729179756.0, "step": 19110 }, { "epoch": 2.431115634143239, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.97240161895752, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8633812665939331, "num_tokens": 729221356.0, "step": 19111 }, { "epoch": 2.4312428444218295, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.996804237365723, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8786174058914185, "num_tokens": 729260456.0, "step": 19112 }, { "epoch": 2.4313700547004196, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.916218757629395, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8619582653045654, "num_tokens": 729299654.0, "step": 19113 }, { "epoch": 2.4314972649790105, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.970511436462402, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8773729801177979, "num_tokens": 729341296.0, "step": 19114 }, { "epoch": 2.4316244752576006, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.954705238342285, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8781583309173584, "num_tokens": 729379632.0, "step": 19115 }, { "epoch": 2.431751685536191, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.01053237915039, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8765777945518494, "num_tokens": 729423202.0, "step": 19116 }, { "epoch": 2.4318788958147817, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.963970184326172, "learning_rate": 1e-06, "loss": 0.3334, "mean_token_accuracy": 0.8937522172927856, "num_tokens": 729457108.0, "step": 19117 }, { "epoch": 2.432006106093372, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.012415885925293, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8707269430160522, "num_tokens": 729494065.0, "step": 19118 }, { "epoch": 2.4321333163719627, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.951468467712402, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8778564929962158, "num_tokens": 729536621.0, "step": 19119 }, { "epoch": 2.4322605266505533, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.969258308410645, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8747937679290771, "num_tokens": 729574423.0, "step": 19120 }, { "epoch": 2.432387736929144, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.936274528503418, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8689227104187012, "num_tokens": 729614673.0, "step": 19121 }, { "epoch": 2.4325149472077343, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.930923461914062, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8759708404541016, "num_tokens": 729653049.0, "step": 19122 }, { "epoch": 2.432642157486325, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.9566068649292, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8748095035552979, "num_tokens": 729689667.0, "step": 19123 }, { "epoch": 2.4327693677649154, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.993908882141113, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8786463141441345, "num_tokens": 729734747.0, "step": 19124 }, { "epoch": 2.432896578043506, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.00204086303711, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.870568037033081, "num_tokens": 729770135.0, "step": 19125 }, { "epoch": 2.4330237883220964, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.988248825073242, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.873184323310852, "num_tokens": 729809203.0, "step": 19126 }, { "epoch": 2.433150998600687, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.95711612701416, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8801664710044861, "num_tokens": 729850164.0, "step": 19127 }, { "epoch": 2.4332782088792775, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.971834182739258, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8639078140258789, "num_tokens": 729884337.0, "step": 19128 }, { "epoch": 2.433405419157868, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.959188461303711, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.887452244758606, "num_tokens": 729924098.0, "step": 19129 }, { "epoch": 2.4335326294364585, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.917947769165039, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8924803137779236, "num_tokens": 729960671.0, "step": 19130 }, { "epoch": 2.433659839715049, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.97256088256836, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8759312629699707, "num_tokens": 730000780.0, "step": 19131 }, { "epoch": 2.4337870499936396, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.972976684570312, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8723005056381226, "num_tokens": 730045198.0, "step": 19132 }, { "epoch": 2.43391426027223, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.02292537689209, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8673343658447266, "num_tokens": 730085010.0, "step": 19133 }, { "epoch": 2.4340414705508207, "ewc_loss": 0.036865234375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 13.912938117980957, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8687783479690552, "num_tokens": 730121512.0, "step": 19134 }, { "epoch": 2.434168680829411, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.979714393615723, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8736989498138428, "num_tokens": 730164859.0, "step": 19135 }, { "epoch": 2.4342958911080017, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.955257415771484, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8742501139640808, "num_tokens": 730202273.0, "step": 19136 }, { "epoch": 2.4344231013865922, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.970854759216309, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.867877721786499, "num_tokens": 730248208.0, "step": 19137 }, { "epoch": 2.4345503116651823, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 14.029726028442383, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8833683729171753, "num_tokens": 730291084.0, "step": 19138 }, { "epoch": 2.4346775219437733, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.944701194763184, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8790668249130249, "num_tokens": 730334504.0, "step": 19139 }, { "epoch": 2.4348047322223634, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.051922798156738, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8781740069389343, "num_tokens": 730373356.0, "step": 19140 }, { "epoch": 2.434931942500954, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.042966842651367, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8617053627967834, "num_tokens": 730416229.0, "step": 19141 }, { "epoch": 2.4350591527795444, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.989544868469238, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.886306881904602, "num_tokens": 730458922.0, "step": 19142 }, { "epoch": 2.435186363058135, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.97817611694336, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8724144697189331, "num_tokens": 730494583.0, "step": 19143 }, { "epoch": 2.4353135733367255, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.009363174438477, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.876774787902832, "num_tokens": 730529526.0, "step": 19144 }, { "epoch": 2.435440783615316, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.986368179321289, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8783281445503235, "num_tokens": 730558584.0, "step": 19145 }, { "epoch": 2.4355679938939065, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 13.863505363464355, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8754563331604004, "num_tokens": 730598677.0, "step": 19146 }, { "epoch": 2.435695204172497, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.056327819824219, "learning_rate": 1e-06, "loss": 0.3661, "mean_token_accuracy": 0.8834362030029297, "num_tokens": 730636985.0, "step": 19147 }, { "epoch": 2.4358224144510876, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.91437816619873, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8568922877311707, "num_tokens": 730675409.0, "step": 19148 }, { "epoch": 2.435949624729678, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.011908531188965, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8626965880393982, "num_tokens": 730714739.0, "step": 19149 }, { "epoch": 2.4360768350082687, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.941438674926758, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.873403787612915, "num_tokens": 730757059.0, "step": 19150 }, { "epoch": 2.436204045286859, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.99039363861084, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8733477592468262, "num_tokens": 730793469.0, "step": 19151 }, { "epoch": 2.4363312555654497, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.966127395629883, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8791009187698364, "num_tokens": 730829336.0, "step": 19152 }, { "epoch": 2.4364584658440402, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.97719669342041, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8727036118507385, "num_tokens": 730869304.0, "step": 19153 }, { "epoch": 2.4365856761226308, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.004223823547363, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.863131046295166, "num_tokens": 730905315.0, "step": 19154 }, { "epoch": 2.4367128864012213, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.967437744140625, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8657429814338684, "num_tokens": 730947346.0, "step": 19155 }, { "epoch": 2.436840096679812, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.986930847167969, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8785484433174133, "num_tokens": 730976907.0, "step": 19156 }, { "epoch": 2.4369673069584024, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.949912071228027, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8704383969306946, "num_tokens": 731015940.0, "step": 19157 }, { "epoch": 2.437094517236993, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.98377799987793, "learning_rate": 1e-06, "loss": 0.3359, "mean_token_accuracy": 0.8903504610061646, "num_tokens": 731051345.0, "step": 19158 }, { "epoch": 2.4372217275155834, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.03896713256836, "learning_rate": 1e-06, "loss": 0.4797, "mean_token_accuracy": 0.8435684442520142, "num_tokens": 731089642.0, "step": 19159 }, { "epoch": 2.437348937794174, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.0111722946167, "learning_rate": 1e-06, "loss": 0.348, "mean_token_accuracy": 0.8907096982002258, "num_tokens": 731133196.0, "step": 19160 }, { "epoch": 2.4374761480727645, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.992965698242188, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8632096648216248, "num_tokens": 731175510.0, "step": 19161 }, { "epoch": 2.437603358351355, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.989861488342285, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.882491946220398, "num_tokens": 731217526.0, "step": 19162 }, { "epoch": 2.437730568629945, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.990994453430176, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8628122210502625, "num_tokens": 731259525.0, "step": 19163 }, { "epoch": 2.437857778908536, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.950148582458496, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8781182765960693, "num_tokens": 731301480.0, "step": 19164 }, { "epoch": 2.437984989187126, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.978647232055664, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8831297159194946, "num_tokens": 731347385.0, "step": 19165 }, { "epoch": 2.4381121994657167, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.986248016357422, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8787049651145935, "num_tokens": 731383666.0, "step": 19166 }, { "epoch": 2.438239409744307, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.96022891998291, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8752484321594238, "num_tokens": 731420297.0, "step": 19167 }, { "epoch": 2.4383666200228977, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.992780685424805, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8707415461540222, "num_tokens": 731455083.0, "step": 19168 }, { "epoch": 2.4384938303014883, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.977453231811523, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.88263338804245, "num_tokens": 731482555.0, "step": 19169 }, { "epoch": 2.438621040580079, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.967948913574219, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8863811492919922, "num_tokens": 731519459.0, "step": 19170 }, { "epoch": 2.4387482508586693, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.881559371948242, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8795580267906189, "num_tokens": 731554577.0, "step": 19171 }, { "epoch": 2.43887546113726, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.022417068481445, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.865617036819458, "num_tokens": 731594993.0, "step": 19172 }, { "epoch": 2.4390026714158504, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.956275939941406, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8785269260406494, "num_tokens": 731629182.0, "step": 19173 }, { "epoch": 2.439129881694441, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.988836288452148, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8729423880577087, "num_tokens": 731667661.0, "step": 19174 }, { "epoch": 2.4392570919730314, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.97510814666748, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8742921948432922, "num_tokens": 731704911.0, "step": 19175 }, { "epoch": 2.439384302251622, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.018683433532715, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8577786684036255, "num_tokens": 731744990.0, "step": 19176 }, { "epoch": 2.4395115125302125, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.9957857131958, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8760976195335388, "num_tokens": 731783335.0, "step": 19177 }, { "epoch": 2.439638722808803, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.017666816711426, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.877321720123291, "num_tokens": 731819206.0, "step": 19178 }, { "epoch": 2.4397659330873935, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.037271499633789, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8722889423370361, "num_tokens": 731855291.0, "step": 19179 }, { "epoch": 2.439893143365984, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.979154586791992, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8782941102981567, "num_tokens": 731899392.0, "step": 19180 }, { "epoch": 2.4400203536445746, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.032946586608887, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8734236359596252, "num_tokens": 731937502.0, "step": 19181 }, { "epoch": 2.440147563923165, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.962600708007812, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8772515058517456, "num_tokens": 731974946.0, "step": 19182 }, { "epoch": 2.4402747742017556, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.985113143920898, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8627774715423584, "num_tokens": 732018732.0, "step": 19183 }, { "epoch": 2.440401984480346, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.985496520996094, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8742349147796631, "num_tokens": 732052036.0, "step": 19184 }, { "epoch": 2.4405291947589367, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.010165214538574, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8697198629379272, "num_tokens": 732090944.0, "step": 19185 }, { "epoch": 2.440656405037527, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.957070350646973, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8766409158706665, "num_tokens": 732131406.0, "step": 19186 }, { "epoch": 2.4407836153161178, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.030631065368652, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8729870915412903, "num_tokens": 732171701.0, "step": 19187 }, { "epoch": 2.440910825594708, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.01947021484375, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8831809163093567, "num_tokens": 732209658.0, "step": 19188 }, { "epoch": 2.4410380358732984, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.009488105773926, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8626234531402588, "num_tokens": 732251395.0, "step": 19189 }, { "epoch": 2.441165246151889, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.99034595489502, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8647022247314453, "num_tokens": 732289306.0, "step": 19190 }, { "epoch": 2.4412924564304794, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.006171226501465, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8598642349243164, "num_tokens": 732337916.0, "step": 19191 }, { "epoch": 2.44141966670907, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.999459266662598, "learning_rate": 1e-06, "loss": 0.4484, "mean_token_accuracy": 0.8609561324119568, "num_tokens": 732380358.0, "step": 19192 }, { "epoch": 2.4415468769876605, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.989114761352539, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8837300539016724, "num_tokens": 732416534.0, "step": 19193 }, { "epoch": 2.441674087266251, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.011415481567383, "learning_rate": 1e-06, "loss": 0.4563, "mean_token_accuracy": 0.8543400764465332, "num_tokens": 732458927.0, "step": 19194 }, { "epoch": 2.4418012975448415, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.95439338684082, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8653541803359985, "num_tokens": 732501261.0, "step": 19195 }, { "epoch": 2.441928507823432, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.96567440032959, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8819393515586853, "num_tokens": 732541162.0, "step": 19196 }, { "epoch": 2.4420557181020226, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.98886775970459, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8715953826904297, "num_tokens": 732578237.0, "step": 19197 }, { "epoch": 2.442182928380613, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.978012084960938, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8636152744293213, "num_tokens": 732618624.0, "step": 19198 }, { "epoch": 2.4423101386592037, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.98735237121582, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8698625564575195, "num_tokens": 732661747.0, "step": 19199 }, { "epoch": 2.442437348937794, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.00737476348877, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8535174131393433, "num_tokens": 732703511.0, "step": 19200 }, { "epoch": 2.4425645592163847, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.917709350585938, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8541807532310486, "num_tokens": 732746749.0, "step": 19201 }, { "epoch": 2.4426917694949752, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.065092086791992, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8757281303405762, "num_tokens": 732783184.0, "step": 19202 }, { "epoch": 2.4428189797735658, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.003954887390137, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8686777353286743, "num_tokens": 732818664.0, "step": 19203 }, { "epoch": 2.4429461900521563, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.060128211975098, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8635442852973938, "num_tokens": 732854286.0, "step": 19204 }, { "epoch": 2.443073400330747, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.992977142333984, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8801438212394714, "num_tokens": 732893032.0, "step": 19205 }, { "epoch": 2.4432006106093374, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.99283218383789, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.873384952545166, "num_tokens": 732929942.0, "step": 19206 }, { "epoch": 2.443327820887928, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.02430534362793, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8696843385696411, "num_tokens": 732968218.0, "step": 19207 }, { "epoch": 2.4434550311665184, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.97183895111084, "learning_rate": 1e-06, "loss": 0.3241, "mean_token_accuracy": 0.8970319032669067, "num_tokens": 733001464.0, "step": 19208 }, { "epoch": 2.443582241445109, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.001138687133789, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8656463623046875, "num_tokens": 733036311.0, "step": 19209 }, { "epoch": 2.4437094517236995, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.964773178100586, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8801454305648804, "num_tokens": 733070487.0, "step": 19210 }, { "epoch": 2.4438366620022896, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.994117736816406, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8713831901550293, "num_tokens": 733110357.0, "step": 19211 }, { "epoch": 2.4439638722808805, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.043924331665039, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8661734461784363, "num_tokens": 733149070.0, "step": 19212 }, { "epoch": 2.4440910825594706, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.050434112548828, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8885648250579834, "num_tokens": 733180848.0, "step": 19213 }, { "epoch": 2.444218292838061, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.99151611328125, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8752495050430298, "num_tokens": 733219204.0, "step": 19214 }, { "epoch": 2.4443455031166517, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.98017406463623, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8634799122810364, "num_tokens": 733255237.0, "step": 19215 }, { "epoch": 2.444472713395242, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.011487007141113, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8849279284477234, "num_tokens": 733293243.0, "step": 19216 }, { "epoch": 2.4445999236738327, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.006036758422852, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8768326640129089, "num_tokens": 733333530.0, "step": 19217 }, { "epoch": 2.4447271339524232, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.03967571258545, "learning_rate": 1e-06, "loss": 0.4548, "mean_token_accuracy": 0.8559092283248901, "num_tokens": 733373790.0, "step": 19218 }, { "epoch": 2.4448543442310138, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.969358444213867, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8769655227661133, "num_tokens": 733416037.0, "step": 19219 }, { "epoch": 2.4449815545096043, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.023869514465332, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.870306670665741, "num_tokens": 733453014.0, "step": 19220 }, { "epoch": 2.445108764788195, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.972575187683105, "learning_rate": 1e-06, "loss": 0.4587, "mean_token_accuracy": 0.8587063550949097, "num_tokens": 733491197.0, "step": 19221 }, { "epoch": 2.4452359750667854, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.010049819946289, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.886272132396698, "num_tokens": 733531142.0, "step": 19222 }, { "epoch": 2.445363185345376, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.001789093017578, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8684241771697998, "num_tokens": 733573916.0, "step": 19223 }, { "epoch": 2.4454903956239664, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.013620376586914, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8683446645736694, "num_tokens": 733613204.0, "step": 19224 }, { "epoch": 2.445617605902557, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.015291213989258, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8687494993209839, "num_tokens": 733653618.0, "step": 19225 }, { "epoch": 2.4457448161811475, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.00639533996582, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8892465233802795, "num_tokens": 733687795.0, "step": 19226 }, { "epoch": 2.445872026459738, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.942706108093262, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.873445987701416, "num_tokens": 733723574.0, "step": 19227 }, { "epoch": 2.4459992367383285, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.99690055847168, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8801113367080688, "num_tokens": 733762051.0, "step": 19228 }, { "epoch": 2.446126447016919, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.986771583557129, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8901441097259521, "num_tokens": 733800640.0, "step": 19229 }, { "epoch": 2.4462536572955096, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.060617446899414, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8632708787918091, "num_tokens": 733841321.0, "step": 19230 }, { "epoch": 2.4463808675741, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.049239158630371, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8766118288040161, "num_tokens": 733874104.0, "step": 19231 }, { "epoch": 2.4465080778526906, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.0018949508667, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8781987428665161, "num_tokens": 733913127.0, "step": 19232 }, { "epoch": 2.446635288131281, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.973297119140625, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.8901795148849487, "num_tokens": 733951900.0, "step": 19233 }, { "epoch": 2.4467624984098717, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.976786613464355, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8757095336914062, "num_tokens": 733992866.0, "step": 19234 }, { "epoch": 2.4468897086884622, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 13.9883451461792, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8686782121658325, "num_tokens": 734029895.0, "step": 19235 }, { "epoch": 2.4470169189670523, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.943819999694824, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8904493451118469, "num_tokens": 734066954.0, "step": 19236 }, { "epoch": 2.4471441292456433, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.986217498779297, "learning_rate": 1e-06, "loss": 0.3336, "mean_token_accuracy": 0.8951180577278137, "num_tokens": 734102412.0, "step": 19237 }, { "epoch": 2.4472713395242334, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.958216667175293, "learning_rate": 1e-06, "loss": 0.3451, "mean_token_accuracy": 0.8908028602600098, "num_tokens": 734141400.0, "step": 19238 }, { "epoch": 2.447398549802824, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.017289161682129, "learning_rate": 1e-06, "loss": 0.4674, "mean_token_accuracy": 0.8541844487190247, "num_tokens": 734176628.0, "step": 19239 }, { "epoch": 2.4475257600814144, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.95300006866455, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8798322677612305, "num_tokens": 734219756.0, "step": 19240 }, { "epoch": 2.447652970360005, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.992547988891602, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8743600845336914, "num_tokens": 734250509.0, "step": 19241 }, { "epoch": 2.4477801806385955, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.957429885864258, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8807367086410522, "num_tokens": 734287887.0, "step": 19242 }, { "epoch": 2.447907390917186, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.002615928649902, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8781331777572632, "num_tokens": 734330446.0, "step": 19243 }, { "epoch": 2.4480346011957765, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.915112495422363, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8790649175643921, "num_tokens": 734368427.0, "step": 19244 }, { "epoch": 2.448161811474367, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.037428855895996, "learning_rate": 1e-06, "loss": 0.3374, "mean_token_accuracy": 0.8930603265762329, "num_tokens": 734405103.0, "step": 19245 }, { "epoch": 2.4482890217529576, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.955939292907715, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8716456294059753, "num_tokens": 734448069.0, "step": 19246 }, { "epoch": 2.448416232031548, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.029547691345215, "learning_rate": 1e-06, "loss": 0.4377, "mean_token_accuracy": 0.8610602617263794, "num_tokens": 734486016.0, "step": 19247 }, { "epoch": 2.4485434423101387, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.988123893737793, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8630537986755371, "num_tokens": 734526943.0, "step": 19248 }, { "epoch": 2.448670652588729, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.994282722473145, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.887070894241333, "num_tokens": 734566137.0, "step": 19249 }, { "epoch": 2.4487978628673197, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.906689643859863, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8718985319137573, "num_tokens": 734602366.0, "step": 19250 }, { "epoch": 2.4489250731459102, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.002692222595215, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8816584348678589, "num_tokens": 734644251.0, "step": 19251 }, { "epoch": 2.4490522834245008, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.953365325927734, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8699400424957275, "num_tokens": 734682938.0, "step": 19252 }, { "epoch": 2.4491794937030913, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.007362365722656, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.875457763671875, "num_tokens": 734722975.0, "step": 19253 }, { "epoch": 2.449306703981682, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.988739967346191, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8774911165237427, "num_tokens": 734761361.0, "step": 19254 }, { "epoch": 2.4494339142602723, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.980963706970215, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8722881078720093, "num_tokens": 734800597.0, "step": 19255 }, { "epoch": 2.449561124538863, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.964123725891113, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8630630373954773, "num_tokens": 734840575.0, "step": 19256 }, { "epoch": 2.4496883348174534, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.972380638122559, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8659913539886475, "num_tokens": 734876943.0, "step": 19257 }, { "epoch": 2.449815545096044, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.994431495666504, "learning_rate": 1e-06, "loss": 0.3485, "mean_token_accuracy": 0.8896197080612183, "num_tokens": 734916436.0, "step": 19258 }, { "epoch": 2.4499427553746345, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.976975440979004, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8730395436286926, "num_tokens": 734956336.0, "step": 19259 }, { "epoch": 2.450069965653225, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.96124267578125, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8729826211929321, "num_tokens": 734997132.0, "step": 19260 }, { "epoch": 2.450197175931815, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.987712860107422, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8625400066375732, "num_tokens": 735036192.0, "step": 19261 }, { "epoch": 2.450324386210406, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.941229820251465, "learning_rate": 1e-06, "loss": 0.3546, "mean_token_accuracy": 0.8870437741279602, "num_tokens": 735079368.0, "step": 19262 }, { "epoch": 2.450451596488996, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.958559036254883, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8800930976867676, "num_tokens": 735109714.0, "step": 19263 }, { "epoch": 2.4505788067675867, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.029722213745117, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.871242105960846, "num_tokens": 735152712.0, "step": 19264 }, { "epoch": 2.450706017046177, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.92022705078125, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8701127767562866, "num_tokens": 735193313.0, "step": 19265 }, { "epoch": 2.4508332273247677, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.014664649963379, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8881022334098816, "num_tokens": 735235413.0, "step": 19266 }, { "epoch": 2.4509604376033582, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.956738471984863, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8756473064422607, "num_tokens": 735270809.0, "step": 19267 }, { "epoch": 2.4510876478819488, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.974804878234863, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.867680013179779, "num_tokens": 735309790.0, "step": 19268 }, { "epoch": 2.4512148581605393, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.00160026550293, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8751555681228638, "num_tokens": 735341084.0, "step": 19269 }, { "epoch": 2.45134206843913, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.985149383544922, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8905156850814819, "num_tokens": 735376964.0, "step": 19270 }, { "epoch": 2.4514692787177204, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.921150207519531, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8837621808052063, "num_tokens": 735413296.0, "step": 19271 }, { "epoch": 2.451596488996311, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.986248970031738, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.878533661365509, "num_tokens": 735454992.0, "step": 19272 }, { "epoch": 2.4517236992749014, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.969326972961426, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.876605749130249, "num_tokens": 735493315.0, "step": 19273 }, { "epoch": 2.451850909553492, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.0176362991333, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8822851777076721, "num_tokens": 735532666.0, "step": 19274 }, { "epoch": 2.4519781198320825, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.004324913024902, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8728986978530884, "num_tokens": 735571987.0, "step": 19275 }, { "epoch": 2.452105330110673, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.035674095153809, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8731123208999634, "num_tokens": 735609409.0, "step": 19276 }, { "epoch": 2.4522325403892635, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.007266998291016, "learning_rate": 1e-06, "loss": 0.4809, "mean_token_accuracy": 0.8484231233596802, "num_tokens": 735647299.0, "step": 19277 }, { "epoch": 2.452359750667854, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.061497688293457, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8828379511833191, "num_tokens": 735677757.0, "step": 19278 }, { "epoch": 2.4524869609464446, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.953704833984375, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.87977534532547, "num_tokens": 735723924.0, "step": 19279 }, { "epoch": 2.452614171225035, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.00186538696289, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8589187860488892, "num_tokens": 735758105.0, "step": 19280 }, { "epoch": 2.4527413815036256, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.049727439880371, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8705896139144897, "num_tokens": 735797603.0, "step": 19281 }, { "epoch": 2.452868591782216, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.007084846496582, "learning_rate": 1e-06, "loss": 0.48, "mean_token_accuracy": 0.8476742506027222, "num_tokens": 735838958.0, "step": 19282 }, { "epoch": 2.4529958020608067, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.972885131835938, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8661404252052307, "num_tokens": 735878894.0, "step": 19283 }, { "epoch": 2.453123012339397, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.040632247924805, "learning_rate": 1e-06, "loss": 0.3079, "mean_token_accuracy": 0.9040963649749756, "num_tokens": 735913222.0, "step": 19284 }, { "epoch": 2.4532502226179878, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.005685806274414, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8812307119369507, "num_tokens": 735951884.0, "step": 19285 }, { "epoch": 2.453377432896578, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.944557189941406, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.869529128074646, "num_tokens": 735998955.0, "step": 19286 }, { "epoch": 2.4535046431751684, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.985329627990723, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8730003833770752, "num_tokens": 736039631.0, "step": 19287 }, { "epoch": 2.453631853453759, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.970202445983887, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8693717122077942, "num_tokens": 736085100.0, "step": 19288 }, { "epoch": 2.4537590637323494, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.049257278442383, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.8896129727363586, "num_tokens": 736119680.0, "step": 19289 }, { "epoch": 2.45388627401094, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.055150985717773, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8864061236381531, "num_tokens": 736159769.0, "step": 19290 }, { "epoch": 2.4540134842895305, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.03320026397705, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8797365427017212, "num_tokens": 736199566.0, "step": 19291 }, { "epoch": 2.454140694568121, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.950108528137207, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8788359761238098, "num_tokens": 736239238.0, "step": 19292 }, { "epoch": 2.4542679048467115, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.053054809570312, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8781028985977173, "num_tokens": 736274576.0, "step": 19293 }, { "epoch": 2.454395115125302, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.983712196350098, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8685300350189209, "num_tokens": 736307509.0, "step": 19294 }, { "epoch": 2.4545223254038926, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.018617630004883, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8702039122581482, "num_tokens": 736348313.0, "step": 19295 }, { "epoch": 2.454649535682483, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.039605140686035, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8707266449928284, "num_tokens": 736380716.0, "step": 19296 }, { "epoch": 2.4547767459610736, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.05081558227539, "learning_rate": 1e-06, "loss": 0.4647, "mean_token_accuracy": 0.8539848327636719, "num_tokens": 736422669.0, "step": 19297 }, { "epoch": 2.454903956239664, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.973858833312988, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.882285475730896, "num_tokens": 736452929.0, "step": 19298 }, { "epoch": 2.4550311665182547, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.068236351013184, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8757394552230835, "num_tokens": 736486755.0, "step": 19299 }, { "epoch": 2.4551583767968452, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.99399471282959, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.882993221282959, "num_tokens": 736521542.0, "step": 19300 }, { "epoch": 2.4552855870754358, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.99409294128418, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8756976127624512, "num_tokens": 736557312.0, "step": 19301 }, { "epoch": 2.4554127973540263, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.007165908813477, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8694790005683899, "num_tokens": 736596469.0, "step": 19302 }, { "epoch": 2.455540007632617, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.006772994995117, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8588185906410217, "num_tokens": 736636078.0, "step": 19303 }, { "epoch": 2.4556672179112073, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.002541542053223, "learning_rate": 1e-06, "loss": 0.3499, "mean_token_accuracy": 0.886688768863678, "num_tokens": 736674117.0, "step": 19304 }, { "epoch": 2.455794428189798, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.071013450622559, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8713943958282471, "num_tokens": 736709116.0, "step": 19305 }, { "epoch": 2.4559216384683884, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.980902671813965, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8659740686416626, "num_tokens": 736750383.0, "step": 19306 }, { "epoch": 2.456048848746979, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.013233184814453, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8634952902793884, "num_tokens": 736790836.0, "step": 19307 }, { "epoch": 2.4561760590255695, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.005851745605469, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8720359802246094, "num_tokens": 736826085.0, "step": 19308 }, { "epoch": 2.4563032693041595, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.059674263000488, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8743258714675903, "num_tokens": 736860599.0, "step": 19309 }, { "epoch": 2.4564304795827505, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.013643264770508, "learning_rate": 1e-06, "loss": 0.336, "mean_token_accuracy": 0.8916354775428772, "num_tokens": 736893286.0, "step": 19310 }, { "epoch": 2.4565576898613406, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.033997535705566, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8841618895530701, "num_tokens": 736931597.0, "step": 19311 }, { "epoch": 2.456684900139931, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.011482238769531, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8828587532043457, "num_tokens": 736972183.0, "step": 19312 }, { "epoch": 2.4568121104185217, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.029877662658691, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8813744783401489, "num_tokens": 737004184.0, "step": 19313 }, { "epoch": 2.456939320697112, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.952264785766602, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8742840886116028, "num_tokens": 737042644.0, "step": 19314 }, { "epoch": 2.4570665309757027, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.968433380126953, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8651061058044434, "num_tokens": 737074830.0, "step": 19315 }, { "epoch": 2.4571937412542932, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.008827209472656, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8825545907020569, "num_tokens": 737107756.0, "step": 19316 }, { "epoch": 2.4573209515328838, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.988774299621582, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8831791281700134, "num_tokens": 737143457.0, "step": 19317 }, { "epoch": 2.4574481618114743, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.024340629577637, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8702641725540161, "num_tokens": 737177005.0, "step": 19318 }, { "epoch": 2.457575372090065, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.969207763671875, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8658807277679443, "num_tokens": 737218265.0, "step": 19319 }, { "epoch": 2.4577025823686554, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.007038116455078, "learning_rate": 1e-06, "loss": 0.4361, "mean_token_accuracy": 0.8643256425857544, "num_tokens": 737259495.0, "step": 19320 }, { "epoch": 2.457829792647246, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.063606262207031, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8669389486312866, "num_tokens": 737300073.0, "step": 19321 }, { "epoch": 2.4579570029258364, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.047515869140625, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8758123517036438, "num_tokens": 737334716.0, "step": 19322 }, { "epoch": 2.458084213204427, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.010512351989746, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.878399670124054, "num_tokens": 737377086.0, "step": 19323 }, { "epoch": 2.4582114234830175, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.026399612426758, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8640424013137817, "num_tokens": 737418259.0, "step": 19324 }, { "epoch": 2.458338633761608, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.041874885559082, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8651366829872131, "num_tokens": 737453150.0, "step": 19325 }, { "epoch": 2.4584658440401985, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.991433143615723, "learning_rate": 1e-06, "loss": 0.3536, "mean_token_accuracy": 0.8895906805992126, "num_tokens": 737495552.0, "step": 19326 }, { "epoch": 2.458593054318789, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.039083480834961, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8799211382865906, "num_tokens": 737537518.0, "step": 19327 }, { "epoch": 2.4587202645973796, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.036005973815918, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8714597821235657, "num_tokens": 737577472.0, "step": 19328 }, { "epoch": 2.45884747487597, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.037286758422852, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8654281497001648, "num_tokens": 737617250.0, "step": 19329 }, { "epoch": 2.4589746851545606, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.999188423156738, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8777332305908203, "num_tokens": 737655544.0, "step": 19330 }, { "epoch": 2.459101895433151, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.0800199508667, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8778966665267944, "num_tokens": 737688644.0, "step": 19331 }, { "epoch": 2.4592291057117417, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.024373054504395, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8780537247657776, "num_tokens": 737730264.0, "step": 19332 }, { "epoch": 2.459356315990332, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.071382522583008, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8689178228378296, "num_tokens": 737763578.0, "step": 19333 }, { "epoch": 2.4594835262689223, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.962204933166504, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8762252330780029, "num_tokens": 737804062.0, "step": 19334 }, { "epoch": 2.4596107365475133, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.016329765319824, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8884108066558838, "num_tokens": 737845608.0, "step": 19335 }, { "epoch": 2.4597379468261034, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.949654579162598, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8573979735374451, "num_tokens": 737881366.0, "step": 19336 }, { "epoch": 2.459865157104694, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.058485984802246, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8675951957702637, "num_tokens": 737922595.0, "step": 19337 }, { "epoch": 2.4599923673832844, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.973169326782227, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8802632093429565, "num_tokens": 737956368.0, "step": 19338 }, { "epoch": 2.460119577661875, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.990107536315918, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8787456750869751, "num_tokens": 738000167.0, "step": 19339 }, { "epoch": 2.4602467879404655, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.084145545959473, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8674346208572388, "num_tokens": 738040850.0, "step": 19340 }, { "epoch": 2.460373998219056, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.073991775512695, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8734956979751587, "num_tokens": 738080370.0, "step": 19341 }, { "epoch": 2.4605012084976465, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.032100677490234, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8799262046813965, "num_tokens": 738116751.0, "step": 19342 }, { "epoch": 2.460628418776237, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.039990425109863, "learning_rate": 1e-06, "loss": 0.3251, "mean_token_accuracy": 0.8920693397521973, "num_tokens": 738154597.0, "step": 19343 }, { "epoch": 2.4607556290548276, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.067011833190918, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8807386159896851, "num_tokens": 738186484.0, "step": 19344 }, { "epoch": 2.460882839333418, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.047600746154785, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8782744407653809, "num_tokens": 738217722.0, "step": 19345 }, { "epoch": 2.4610100496120086, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.079703330993652, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8878421783447266, "num_tokens": 738255304.0, "step": 19346 }, { "epoch": 2.461137259890599, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.96750545501709, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8701624274253845, "num_tokens": 738292162.0, "step": 19347 }, { "epoch": 2.4612644701691897, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.07893180847168, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8859505653381348, "num_tokens": 738332851.0, "step": 19348 }, { "epoch": 2.4613916804477802, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.958187103271484, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8854060769081116, "num_tokens": 738375120.0, "step": 19349 }, { "epoch": 2.4615188907263708, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.034602165222168, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8659122586250305, "num_tokens": 738413857.0, "step": 19350 }, { "epoch": 2.4616461010049613, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.012468338012695, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8688108921051025, "num_tokens": 738450097.0, "step": 19351 }, { "epoch": 2.461773311283552, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.097091674804688, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8567225337028503, "num_tokens": 738483784.0, "step": 19352 }, { "epoch": 2.4619005215621423, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.99267578125, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8555892705917358, "num_tokens": 738526820.0, "step": 19353 }, { "epoch": 2.462027731840733, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.065824508666992, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8608609437942505, "num_tokens": 738569049.0, "step": 19354 }, { "epoch": 2.4621549421193234, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.037261962890625, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8709208965301514, "num_tokens": 738608320.0, "step": 19355 }, { "epoch": 2.462282152397914, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.017692565917969, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8892369270324707, "num_tokens": 738645566.0, "step": 19356 }, { "epoch": 2.4624093626765045, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.053853988647461, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8524048924446106, "num_tokens": 738686184.0, "step": 19357 }, { "epoch": 2.462536572955095, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.009759902954102, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8611652851104736, "num_tokens": 738728820.0, "step": 19358 }, { "epoch": 2.462663783233685, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.08161449432373, "learning_rate": 1e-06, "loss": 0.487, "mean_token_accuracy": 0.8457035422325134, "num_tokens": 738760180.0, "step": 19359 }, { "epoch": 2.462790993512276, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.046459197998047, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8829124569892883, "num_tokens": 738795501.0, "step": 19360 }, { "epoch": 2.462918203790866, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.040237426757812, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8651291131973267, "num_tokens": 738831027.0, "step": 19361 }, { "epoch": 2.4630454140694567, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.031691551208496, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8573610186576843, "num_tokens": 738870383.0, "step": 19362 }, { "epoch": 2.463172624348047, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.010242462158203, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8786656856536865, "num_tokens": 738908231.0, "step": 19363 }, { "epoch": 2.4632998346266377, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.11074447631836, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8655937910079956, "num_tokens": 738950927.0, "step": 19364 }, { "epoch": 2.4634270449052282, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.009821891784668, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8860812783241272, "num_tokens": 738989857.0, "step": 19365 }, { "epoch": 2.4635542551838188, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.107297897338867, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8877825736999512, "num_tokens": 739021281.0, "step": 19366 }, { "epoch": 2.4636814654624093, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.053505897521973, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8816472887992859, "num_tokens": 739063475.0, "step": 19367 }, { "epoch": 2.463808675741, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.088759422302246, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.870154857635498, "num_tokens": 739107083.0, "step": 19368 }, { "epoch": 2.4639358860195903, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.042717933654785, "learning_rate": 1e-06, "loss": 0.4605, "mean_token_accuracy": 0.8534488677978516, "num_tokens": 739152002.0, "step": 19369 }, { "epoch": 2.464063096298181, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.073495864868164, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8717162609100342, "num_tokens": 739191781.0, "step": 19370 }, { "epoch": 2.4641903065767714, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.050134658813477, "learning_rate": 1e-06, "loss": 0.3388, "mean_token_accuracy": 0.8941640257835388, "num_tokens": 739230374.0, "step": 19371 }, { "epoch": 2.464317516855362, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.990768432617188, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8713250160217285, "num_tokens": 739267310.0, "step": 19372 }, { "epoch": 2.4644447271339525, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.113021850585938, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8772181868553162, "num_tokens": 739305074.0, "step": 19373 }, { "epoch": 2.464571937412543, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.088665962219238, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8665728569030762, "num_tokens": 739343028.0, "step": 19374 }, { "epoch": 2.4646991476911335, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.01453685760498, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8732447028160095, "num_tokens": 739384364.0, "step": 19375 }, { "epoch": 2.464826357969724, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.107094764709473, "learning_rate": 1e-06, "loss": 0.3493, "mean_token_accuracy": 0.8899668455123901, "num_tokens": 739423564.0, "step": 19376 }, { "epoch": 2.4649535682483146, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.043862342834473, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8820788264274597, "num_tokens": 739457609.0, "step": 19377 }, { "epoch": 2.465080778526905, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.033875465393066, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8753707408905029, "num_tokens": 739497204.0, "step": 19378 }, { "epoch": 2.4652079888054956, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.057259559631348, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8680605888366699, "num_tokens": 739543688.0, "step": 19379 }, { "epoch": 2.465335199084086, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.092642784118652, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8856534957885742, "num_tokens": 739578221.0, "step": 19380 }, { "epoch": 2.4654624093626767, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.032570838928223, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.880038321018219, "num_tokens": 739615892.0, "step": 19381 }, { "epoch": 2.4655896196412668, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.045470237731934, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8601231575012207, "num_tokens": 739657697.0, "step": 19382 }, { "epoch": 2.4657168299198577, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.053367614746094, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8830827474594116, "num_tokens": 739694173.0, "step": 19383 }, { "epoch": 2.465844040198448, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.01626968383789, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8716220259666443, "num_tokens": 739734111.0, "step": 19384 }, { "epoch": 2.4659712504770384, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.12291431427002, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8755948543548584, "num_tokens": 739771409.0, "step": 19385 }, { "epoch": 2.466098460755629, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.032818794250488, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8611884117126465, "num_tokens": 739807185.0, "step": 19386 }, { "epoch": 2.4662256710342194, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.00969123840332, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8868623971939087, "num_tokens": 739837750.0, "step": 19387 }, { "epoch": 2.46635288131281, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.061863899230957, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8734067678451538, "num_tokens": 739879516.0, "step": 19388 }, { "epoch": 2.4664800915914005, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.037471771240234, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8809168934822083, "num_tokens": 739918688.0, "step": 19389 }, { "epoch": 2.466607301869991, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.08065414428711, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8787484765052795, "num_tokens": 739958978.0, "step": 19390 }, { "epoch": 2.4667345121485815, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.015340805053711, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8790092468261719, "num_tokens": 740002428.0, "step": 19391 }, { "epoch": 2.466861722427172, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.051342010498047, "learning_rate": 1e-06, "loss": 0.3424, "mean_token_accuracy": 0.894149899482727, "num_tokens": 740045856.0, "step": 19392 }, { "epoch": 2.4669889327057626, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.043275833129883, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8717691898345947, "num_tokens": 740084512.0, "step": 19393 }, { "epoch": 2.467116142984353, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.979693412780762, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8846598267555237, "num_tokens": 740129624.0, "step": 19394 }, { "epoch": 2.4672433532629436, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.025328636169434, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8583427667617798, "num_tokens": 740168368.0, "step": 19395 }, { "epoch": 2.467370563541534, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.07022762298584, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8679976463317871, "num_tokens": 740206670.0, "step": 19396 }, { "epoch": 2.4674977738201247, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.021364212036133, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8520046472549438, "num_tokens": 740246422.0, "step": 19397 }, { "epoch": 2.4676249840987152, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.024381637573242, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8713982701301575, "num_tokens": 740284450.0, "step": 19398 }, { "epoch": 2.4677521943773058, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.003570556640625, "learning_rate": 1e-06, "loss": 0.4611, "mean_token_accuracy": 0.8517370223999023, "num_tokens": 740319062.0, "step": 19399 }, { "epoch": 2.4678794046558963, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.10201644897461, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8752671480178833, "num_tokens": 740353996.0, "step": 19400 }, { "epoch": 2.468006614934487, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.023615837097168, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8726493120193481, "num_tokens": 740394694.0, "step": 19401 }, { "epoch": 2.4681338252130773, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.001822471618652, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.8938419818878174, "num_tokens": 740436567.0, "step": 19402 }, { "epoch": 2.468261035491668, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.993036270141602, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8666956424713135, "num_tokens": 740476458.0, "step": 19403 }, { "epoch": 2.4683882457702584, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.09666919708252, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8822798728942871, "num_tokens": 740514703.0, "step": 19404 }, { "epoch": 2.468515456048849, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.020289421081543, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8844715356826782, "num_tokens": 740552328.0, "step": 19405 }, { "epoch": 2.4686426663274394, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.069794654846191, "learning_rate": 1e-06, "loss": 0.4499, "mean_token_accuracy": 0.8562361001968384, "num_tokens": 740593141.0, "step": 19406 }, { "epoch": 2.4687698766060295, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.973516464233398, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8843199014663696, "num_tokens": 740634097.0, "step": 19407 }, { "epoch": 2.4688970868846205, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.003021240234375, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8793702125549316, "num_tokens": 740672917.0, "step": 19408 }, { "epoch": 2.4690242971632106, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.063028335571289, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8697370290756226, "num_tokens": 740709351.0, "step": 19409 }, { "epoch": 2.469151507441801, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.018234252929688, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8645685315132141, "num_tokens": 740744049.0, "step": 19410 }, { "epoch": 2.4692787177203916, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.016651153564453, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8770778179168701, "num_tokens": 740784765.0, "step": 19411 }, { "epoch": 2.469405927998982, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.119880676269531, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8674795031547546, "num_tokens": 740815859.0, "step": 19412 }, { "epoch": 2.4695331382775727, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.012805938720703, "learning_rate": 1e-06, "loss": 0.3366, "mean_token_accuracy": 0.8916299939155579, "num_tokens": 740854300.0, "step": 19413 }, { "epoch": 2.4696603485561632, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.987133026123047, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8579115271568298, "num_tokens": 740891836.0, "step": 19414 }, { "epoch": 2.4697875588347538, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.021158218383789, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8830434679985046, "num_tokens": 740927434.0, "step": 19415 }, { "epoch": 2.4699147691133443, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.00114631652832, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8763120174407959, "num_tokens": 740973707.0, "step": 19416 }, { "epoch": 2.470041979391935, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.082756996154785, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8813229203224182, "num_tokens": 741015953.0, "step": 19417 }, { "epoch": 2.4701691896705253, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.0890474319458, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8809739947319031, "num_tokens": 741049634.0, "step": 19418 }, { "epoch": 2.470296399949116, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.991347312927246, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8841186761856079, "num_tokens": 741084267.0, "step": 19419 }, { "epoch": 2.4704236102277064, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.003551483154297, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8798649907112122, "num_tokens": 741122237.0, "step": 19420 }, { "epoch": 2.470550820506297, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.064050674438477, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8624818921089172, "num_tokens": 741163593.0, "step": 19421 }, { "epoch": 2.4706780307848875, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.030108451843262, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8641030788421631, "num_tokens": 741202328.0, "step": 19422 }, { "epoch": 2.470805241063478, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.080735206604004, "learning_rate": 1e-06, "loss": 0.3355, "mean_token_accuracy": 0.893548846244812, "num_tokens": 741240095.0, "step": 19423 }, { "epoch": 2.4709324513420685, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.964229583740234, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8760628700256348, "num_tokens": 741278509.0, "step": 19424 }, { "epoch": 2.471059661620659, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.119629859924316, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8695160150527954, "num_tokens": 741321446.0, "step": 19425 }, { "epoch": 2.4711868718992496, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.006903648376465, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8702067732810974, "num_tokens": 741360469.0, "step": 19426 }, { "epoch": 2.47131408217784, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.06856632232666, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.865203320980072, "num_tokens": 741393833.0, "step": 19427 }, { "epoch": 2.4714412924564306, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.07125473022461, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.879186749458313, "num_tokens": 741426956.0, "step": 19428 }, { "epoch": 2.471568502735021, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.003926277160645, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8730671405792236, "num_tokens": 741464820.0, "step": 19429 }, { "epoch": 2.4716957130136117, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.071861267089844, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8811074495315552, "num_tokens": 741499519.0, "step": 19430 }, { "epoch": 2.471822923292202, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.00137996673584, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.881459653377533, "num_tokens": 741533443.0, "step": 19431 }, { "epoch": 2.4719501335707923, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.040140151977539, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.870911717414856, "num_tokens": 741576849.0, "step": 19432 }, { "epoch": 2.4720773438493833, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.996570587158203, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8795951008796692, "num_tokens": 741614653.0, "step": 19433 }, { "epoch": 2.4722045541279734, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.037076950073242, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8692018985748291, "num_tokens": 741649354.0, "step": 19434 }, { "epoch": 2.472331764406564, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.010251998901367, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8668006062507629, "num_tokens": 741692257.0, "step": 19435 }, { "epoch": 2.4724589746851544, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.039965629577637, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8718529939651489, "num_tokens": 741729739.0, "step": 19436 }, { "epoch": 2.472586184963745, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.034784317016602, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8646960258483887, "num_tokens": 741763409.0, "step": 19437 }, { "epoch": 2.4727133952423355, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.98945426940918, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8709896206855774, "num_tokens": 741798746.0, "step": 19438 }, { "epoch": 2.472840605520926, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.000247955322266, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8829920291900635, "num_tokens": 741840488.0, "step": 19439 }, { "epoch": 2.4729678157995165, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.972160339355469, "learning_rate": 1e-06, "loss": 0.3461, "mean_token_accuracy": 0.8904627561569214, "num_tokens": 741880839.0, "step": 19440 }, { "epoch": 2.473095026078107, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.042194366455078, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8682348132133484, "num_tokens": 741919317.0, "step": 19441 }, { "epoch": 2.4732222363566976, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.023455619812012, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8550824522972107, "num_tokens": 741955782.0, "step": 19442 }, { "epoch": 2.473349446635288, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.01931381225586, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8684202432632446, "num_tokens": 741990410.0, "step": 19443 }, { "epoch": 2.4734766569138786, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.033985137939453, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.869573712348938, "num_tokens": 742031536.0, "step": 19444 }, { "epoch": 2.473603867192469, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.090384483337402, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8644919395446777, "num_tokens": 742067222.0, "step": 19445 }, { "epoch": 2.4737310774710597, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.03258991241455, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.885483980178833, "num_tokens": 742114740.0, "step": 19446 }, { "epoch": 2.47385828774965, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.067854881286621, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8832352757453918, "num_tokens": 742149344.0, "step": 19447 }, { "epoch": 2.4739854980282407, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.03242301940918, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8761863708496094, "num_tokens": 742191089.0, "step": 19448 }, { "epoch": 2.4741127083068313, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.121476173400879, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8746886253356934, "num_tokens": 742224944.0, "step": 19449 }, { "epoch": 2.474239918585422, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.013825416564941, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8637371063232422, "num_tokens": 742263674.0, "step": 19450 }, { "epoch": 2.4743671288640123, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.095404624938965, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8684302568435669, "num_tokens": 742300424.0, "step": 19451 }, { "epoch": 2.474494339142603, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.056796073913574, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8835038542747498, "num_tokens": 742333854.0, "step": 19452 }, { "epoch": 2.4746215494211934, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.017837524414062, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8819133043289185, "num_tokens": 742377126.0, "step": 19453 }, { "epoch": 2.474748759699784, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.077200889587402, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8720289468765259, "num_tokens": 742421284.0, "step": 19454 }, { "epoch": 2.4748759699783744, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.064006805419922, "learning_rate": 1e-06, "loss": 0.4594, "mean_token_accuracy": 0.8575559258460999, "num_tokens": 742459124.0, "step": 19455 }, { "epoch": 2.475003180256965, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.0761137008667, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8649190664291382, "num_tokens": 742493137.0, "step": 19456 }, { "epoch": 2.475130390535555, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.09781551361084, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8768101334571838, "num_tokens": 742529805.0, "step": 19457 }, { "epoch": 2.475257600814146, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.039803504943848, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8871598243713379, "num_tokens": 742560859.0, "step": 19458 }, { "epoch": 2.475384811092736, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 14.007878303527832, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8856187462806702, "num_tokens": 742601465.0, "step": 19459 }, { "epoch": 2.4755120213713266, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.002787590026855, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8678181171417236, "num_tokens": 742639242.0, "step": 19460 }, { "epoch": 2.475639231649917, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.07369613647461, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8739949464797974, "num_tokens": 742676574.0, "step": 19461 }, { "epoch": 2.4757664419285077, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.015186309814453, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8640573024749756, "num_tokens": 742713326.0, "step": 19462 }, { "epoch": 2.4758936522070982, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.082880020141602, "learning_rate": 1e-06, "loss": 0.461, "mean_token_accuracy": 0.8511468768119812, "num_tokens": 742756883.0, "step": 19463 }, { "epoch": 2.4760208624856888, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.060691833496094, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8695595264434814, "num_tokens": 742794301.0, "step": 19464 }, { "epoch": 2.4761480727642793, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.9933500289917, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8763197660446167, "num_tokens": 742832966.0, "step": 19465 }, { "epoch": 2.47627528304287, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.060378074645996, "learning_rate": 1e-06, "loss": 0.4204, "mean_token_accuracy": 0.8702394962310791, "num_tokens": 742872926.0, "step": 19466 }, { "epoch": 2.4764024933214603, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.059514045715332, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8818437457084656, "num_tokens": 742910838.0, "step": 19467 }, { "epoch": 2.476529703600051, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.968008995056152, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8821979761123657, "num_tokens": 742950098.0, "step": 19468 }, { "epoch": 2.4766569138786414, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.043582916259766, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8842787742614746, "num_tokens": 742983872.0, "step": 19469 }, { "epoch": 2.476784124157232, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.97121810913086, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.863818883895874, "num_tokens": 743029816.0, "step": 19470 }, { "epoch": 2.4769113344358225, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.057442665100098, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8664965629577637, "num_tokens": 743068565.0, "step": 19471 }, { "epoch": 2.477038544714413, "ewc_loss": 0.037109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 13.932790756225586, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8756139278411865, "num_tokens": 743112137.0, "step": 19472 }, { "epoch": 2.4771657549930035, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.018588066101074, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8891241550445557, "num_tokens": 743145067.0, "step": 19473 }, { "epoch": 2.477292965271594, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.993086814880371, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8786938786506653, "num_tokens": 743177017.0, "step": 19474 }, { "epoch": 2.4774201755501846, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.007657051086426, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8741111755371094, "num_tokens": 743218121.0, "step": 19475 }, { "epoch": 2.477547385828775, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.981176376342773, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8798913955688477, "num_tokens": 743256635.0, "step": 19476 }, { "epoch": 2.4776745961073656, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.054006576538086, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8721315860748291, "num_tokens": 743293801.0, "step": 19477 }, { "epoch": 2.477801806385956, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.00649356842041, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8873460292816162, "num_tokens": 743331582.0, "step": 19478 }, { "epoch": 2.4779290166645467, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.01601791381836, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8768383264541626, "num_tokens": 743369901.0, "step": 19479 }, { "epoch": 2.4780562269431368, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.06186580657959, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8697084784507751, "num_tokens": 743405808.0, "step": 19480 }, { "epoch": 2.4781834372217277, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.991209983825684, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8759618997573853, "num_tokens": 743447236.0, "step": 19481 }, { "epoch": 2.478310647500318, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.033210754394531, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8814338445663452, "num_tokens": 743486494.0, "step": 19482 }, { "epoch": 2.4784378577789083, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.036800384521484, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8703840970993042, "num_tokens": 743523941.0, "step": 19483 }, { "epoch": 2.478565068057499, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.036914825439453, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8786051273345947, "num_tokens": 743554169.0, "step": 19484 }, { "epoch": 2.4786922783360894, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.057449340820312, "learning_rate": 1e-06, "loss": 0.4483, "mean_token_accuracy": 0.8584194779396057, "num_tokens": 743594656.0, "step": 19485 }, { "epoch": 2.47881948861468, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.057519912719727, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8802330493927002, "num_tokens": 743628638.0, "step": 19486 }, { "epoch": 2.4789466988932705, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.007794380187988, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.866094708442688, "num_tokens": 743665382.0, "step": 19487 }, { "epoch": 2.479073909171861, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.015496253967285, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8867676258087158, "num_tokens": 743701712.0, "step": 19488 }, { "epoch": 2.4792011194504515, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.967939376831055, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8745367527008057, "num_tokens": 743740347.0, "step": 19489 }, { "epoch": 2.479328329729042, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.992207527160645, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8764336705207825, "num_tokens": 743775434.0, "step": 19490 }, { "epoch": 2.4794555400076326, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.084792137145996, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8740061521530151, "num_tokens": 743812475.0, "step": 19491 }, { "epoch": 2.479582750286223, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.940681457519531, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8656731247901917, "num_tokens": 743851350.0, "step": 19492 }, { "epoch": 2.4797099605648136, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.026341438293457, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8722308874130249, "num_tokens": 743885086.0, "step": 19493 }, { "epoch": 2.479837170843404, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 13.983808517456055, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8765317797660828, "num_tokens": 743931747.0, "step": 19494 }, { "epoch": 2.4799643811219947, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.032735824584961, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8872051239013672, "num_tokens": 743971806.0, "step": 19495 }, { "epoch": 2.480091591400585, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.040435791015625, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8843036890029907, "num_tokens": 744005304.0, "step": 19496 }, { "epoch": 2.4802188016791757, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.007246971130371, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.865612268447876, "num_tokens": 744051159.0, "step": 19497 }, { "epoch": 2.4803460119577663, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.049080848693848, "learning_rate": 1e-06, "loss": 0.4466, "mean_token_accuracy": 0.8561869263648987, "num_tokens": 744089856.0, "step": 19498 }, { "epoch": 2.480473222236357, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.042563438415527, "learning_rate": 1e-06, "loss": 0.4659, "mean_token_accuracy": 0.8550142049789429, "num_tokens": 744136355.0, "step": 19499 }, { "epoch": 2.4806004325149473, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.03199577331543, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8637543320655823, "num_tokens": 744177682.0, "step": 19500 }, { "epoch": 2.480727642793538, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.091390609741211, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.8749686479568481, "num_tokens": 744214209.0, "step": 19501 }, { "epoch": 2.4808548530721284, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.043169021606445, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8571437001228333, "num_tokens": 744257923.0, "step": 19502 }, { "epoch": 2.480982063350719, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.059521675109863, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8808218240737915, "num_tokens": 744293370.0, "step": 19503 }, { "epoch": 2.4811092736293094, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.118749618530273, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.880347490310669, "num_tokens": 744328572.0, "step": 19504 }, { "epoch": 2.4812364839078995, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.02990436553955, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8581324219703674, "num_tokens": 744364124.0, "step": 19505 }, { "epoch": 2.4813636941864905, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.110108375549316, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8771607875823975, "num_tokens": 744403679.0, "step": 19506 }, { "epoch": 2.4814909044650806, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.073622703552246, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.874199628829956, "num_tokens": 744440922.0, "step": 19507 }, { "epoch": 2.481618114743671, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.020576477050781, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8720312118530273, "num_tokens": 744479025.0, "step": 19508 }, { "epoch": 2.4817453250222616, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.094453811645508, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8759229183197021, "num_tokens": 744514165.0, "step": 19509 }, { "epoch": 2.481872535300852, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.055902481079102, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.86906498670578, "num_tokens": 744547210.0, "step": 19510 }, { "epoch": 2.4819997455794427, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.076778411865234, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8736975193023682, "num_tokens": 744585132.0, "step": 19511 }, { "epoch": 2.4821269558580332, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.051481246948242, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8847248554229736, "num_tokens": 744617921.0, "step": 19512 }, { "epoch": 2.4822541661366238, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.02252197265625, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8688011765480042, "num_tokens": 744653095.0, "step": 19513 }, { "epoch": 2.4823813764152143, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.058106422424316, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8800297975540161, "num_tokens": 744688258.0, "step": 19514 }, { "epoch": 2.482508586693805, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.058340072631836, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8735822439193726, "num_tokens": 744723500.0, "step": 19515 }, { "epoch": 2.4826357969723953, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.10917854309082, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8750384449958801, "num_tokens": 744757133.0, "step": 19516 }, { "epoch": 2.482763007250986, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.047608375549316, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8868275880813599, "num_tokens": 744793814.0, "step": 19517 }, { "epoch": 2.4828902175295764, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.009541511535645, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8776397705078125, "num_tokens": 744828680.0, "step": 19518 }, { "epoch": 2.483017427808167, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.095518112182617, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8685294389724731, "num_tokens": 744861642.0, "step": 19519 }, { "epoch": 2.4831446380867574, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.069804191589355, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8849778175354004, "num_tokens": 744900345.0, "step": 19520 }, { "epoch": 2.483271848365348, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.009090423583984, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8805514574050903, "num_tokens": 744938589.0, "step": 19521 }, { "epoch": 2.4833990586439385, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.016685485839844, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8654220104217529, "num_tokens": 744975385.0, "step": 19522 }, { "epoch": 2.483526268922529, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.061495780944824, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8641360998153687, "num_tokens": 745010860.0, "step": 19523 }, { "epoch": 2.4836534792011196, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.04331111907959, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8661568760871887, "num_tokens": 745047183.0, "step": 19524 }, { "epoch": 2.48378068947971, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.022124290466309, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8714816570281982, "num_tokens": 745081906.0, "step": 19525 }, { "epoch": 2.4839078997583006, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.990801811218262, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8785871267318726, "num_tokens": 745127285.0, "step": 19526 }, { "epoch": 2.484035110036891, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.022066116333008, "learning_rate": 1e-06, "loss": 0.352, "mean_token_accuracy": 0.8900488615036011, "num_tokens": 745171564.0, "step": 19527 }, { "epoch": 2.4841623203154817, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.065559387207031, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8846012353897095, "num_tokens": 745214252.0, "step": 19528 }, { "epoch": 2.484289530594072, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.971097946166992, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8767403960227966, "num_tokens": 745251983.0, "step": 19529 }, { "epoch": 2.4844167408726623, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.041903495788574, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8760957717895508, "num_tokens": 745287037.0, "step": 19530 }, { "epoch": 2.4845439511512533, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.050860404968262, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8710405826568604, "num_tokens": 745327854.0, "step": 19531 }, { "epoch": 2.4846711614298433, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 13.964054107666016, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8624573945999146, "num_tokens": 745370227.0, "step": 19532 }, { "epoch": 2.484798371708434, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.010327339172363, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8718399405479431, "num_tokens": 745409416.0, "step": 19533 }, { "epoch": 2.4849255819870244, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.02502727508545, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.857002854347229, "num_tokens": 745449908.0, "step": 19534 }, { "epoch": 2.485052792265615, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.003132820129395, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8636921048164368, "num_tokens": 745486384.0, "step": 19535 }, { "epoch": 2.4851800025442055, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.053984642028809, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.862729549407959, "num_tokens": 745523058.0, "step": 19536 }, { "epoch": 2.485307212822796, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.059687614440918, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8568199276924133, "num_tokens": 745559276.0, "step": 19537 }, { "epoch": 2.4854344231013865, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.07848834991455, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8721665143966675, "num_tokens": 745592726.0, "step": 19538 }, { "epoch": 2.485561633379977, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.028156280517578, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8750447630882263, "num_tokens": 745630117.0, "step": 19539 }, { "epoch": 2.4856888436585676, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.065847396850586, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8599926233291626, "num_tokens": 745664968.0, "step": 19540 }, { "epoch": 2.485816053937158, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.069937705993652, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8751312494277954, "num_tokens": 745701893.0, "step": 19541 }, { "epoch": 2.4859432642157486, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.020227432250977, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8703662157058716, "num_tokens": 745736884.0, "step": 19542 }, { "epoch": 2.486070474494339, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.059174537658691, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8851858377456665, "num_tokens": 745782381.0, "step": 19543 }, { "epoch": 2.4861976847729297, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.00167465209961, "learning_rate": 1e-06, "loss": 0.3487, "mean_token_accuracy": 0.8893911838531494, "num_tokens": 745827630.0, "step": 19544 }, { "epoch": 2.48632489505152, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.110482215881348, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8815606236457825, "num_tokens": 745862223.0, "step": 19545 }, { "epoch": 2.4864521053301107, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.072271347045898, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8890680074691772, "num_tokens": 745903077.0, "step": 19546 }, { "epoch": 2.4865793156087013, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.073333740234375, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8737550973892212, "num_tokens": 745945534.0, "step": 19547 }, { "epoch": 2.486706525887292, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.11818790435791, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8575723171234131, "num_tokens": 745986308.0, "step": 19548 }, { "epoch": 2.4868337361658823, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.087775230407715, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8651262521743774, "num_tokens": 746022206.0, "step": 19549 }, { "epoch": 2.486960946444473, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.057170867919922, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.880347728729248, "num_tokens": 746059854.0, "step": 19550 }, { "epoch": 2.4870881567230634, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.033809661865234, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.881921648979187, "num_tokens": 746103994.0, "step": 19551 }, { "epoch": 2.487215367001654, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.048990249633789, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8640249967575073, "num_tokens": 746138491.0, "step": 19552 }, { "epoch": 2.4873425772802444, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.081995010375977, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8723238706588745, "num_tokens": 746172598.0, "step": 19553 }, { "epoch": 2.487469787558835, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.07250690460205, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8800341486930847, "num_tokens": 746210385.0, "step": 19554 }, { "epoch": 2.487596997837425, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.013943672180176, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8756699562072754, "num_tokens": 746252563.0, "step": 19555 }, { "epoch": 2.487724208116016, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.041007041931152, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8787071704864502, "num_tokens": 746292356.0, "step": 19556 }, { "epoch": 2.487851418394606, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.033702850341797, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8638351559638977, "num_tokens": 746325467.0, "step": 19557 }, { "epoch": 2.4879786286731966, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.07215690612793, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8694754838943481, "num_tokens": 746362285.0, "step": 19558 }, { "epoch": 2.488105838951787, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.991302490234375, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8803352117538452, "num_tokens": 746401540.0, "step": 19559 }, { "epoch": 2.4882330492303777, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.041423797607422, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8754174709320068, "num_tokens": 746439282.0, "step": 19560 }, { "epoch": 2.488360259508968, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.018827438354492, "learning_rate": 1e-06, "loss": 0.3663, "mean_token_accuracy": 0.8847569227218628, "num_tokens": 746480993.0, "step": 19561 }, { "epoch": 2.4884874697875587, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.050602912902832, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8713697195053101, "num_tokens": 746519318.0, "step": 19562 }, { "epoch": 2.4886146800661493, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.034159660339355, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8611252307891846, "num_tokens": 746559761.0, "step": 19563 }, { "epoch": 2.48874189034474, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.073385238647461, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.881309449672699, "num_tokens": 746591622.0, "step": 19564 }, { "epoch": 2.4888691006233303, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.050219535827637, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8724819421768188, "num_tokens": 746627709.0, "step": 19565 }, { "epoch": 2.488996310901921, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.071279525756836, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8706562519073486, "num_tokens": 746663445.0, "step": 19566 }, { "epoch": 2.4891235211805114, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.032671928405762, "learning_rate": 1e-06, "loss": 0.4295, "mean_token_accuracy": 0.8625912666320801, "num_tokens": 746706869.0, "step": 19567 }, { "epoch": 2.489250731459102, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.080551147460938, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8861251473426819, "num_tokens": 746749502.0, "step": 19568 }, { "epoch": 2.4893779417376924, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.083517074584961, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8761313557624817, "num_tokens": 746786510.0, "step": 19569 }, { "epoch": 2.489505152016283, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.049965858459473, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.879448413848877, "num_tokens": 746819430.0, "step": 19570 }, { "epoch": 2.4896323622948735, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.058969497680664, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.877109169960022, "num_tokens": 746856018.0, "step": 19571 }, { "epoch": 2.489759572573464, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.064624786376953, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8684892654418945, "num_tokens": 746897350.0, "step": 19572 }, { "epoch": 2.4898867828520546, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.04802417755127, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8679522275924683, "num_tokens": 746937163.0, "step": 19573 }, { "epoch": 2.490013993130645, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.043706893920898, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8845151662826538, "num_tokens": 746976140.0, "step": 19574 }, { "epoch": 2.4901412034092356, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.082477569580078, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8733015656471252, "num_tokens": 747012576.0, "step": 19575 }, { "epoch": 2.490268413687826, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 13.995369911193848, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8722942471504211, "num_tokens": 747048179.0, "step": 19576 }, { "epoch": 2.4903956239664167, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.105510711669922, "learning_rate": 1e-06, "loss": 0.4492, "mean_token_accuracy": 0.8546355366706848, "num_tokens": 747085291.0, "step": 19577 }, { "epoch": 2.4905228342450068, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.05997085571289, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8815668225288391, "num_tokens": 747119085.0, "step": 19578 }, { "epoch": 2.4906500445235977, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.040945053100586, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8715783357620239, "num_tokens": 747153463.0, "step": 19579 }, { "epoch": 2.490777254802188, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.079019546508789, "learning_rate": 1e-06, "loss": 0.4778, "mean_token_accuracy": 0.8495807647705078, "num_tokens": 747194781.0, "step": 19580 }, { "epoch": 2.4909044650807783, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.9948091506958, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8757128715515137, "num_tokens": 747235328.0, "step": 19581 }, { "epoch": 2.491031675359369, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.055207252502441, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8722442984580994, "num_tokens": 747278481.0, "step": 19582 }, { "epoch": 2.4911588856379594, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 13.976771354675293, "learning_rate": 1e-06, "loss": 0.3517, "mean_token_accuracy": 0.8874996900558472, "num_tokens": 747313077.0, "step": 19583 }, { "epoch": 2.49128609591655, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.031805038452148, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8808586597442627, "num_tokens": 747352693.0, "step": 19584 }, { "epoch": 2.4914133061951405, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 13.976826667785645, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8659148216247559, "num_tokens": 747397717.0, "step": 19585 }, { "epoch": 2.491540516473731, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.042588233947754, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8770827054977417, "num_tokens": 747435457.0, "step": 19586 }, { "epoch": 2.4916677267523215, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.022401809692383, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8781498074531555, "num_tokens": 747479017.0, "step": 19587 }, { "epoch": 2.491794937030912, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.061149597167969, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8737848997116089, "num_tokens": 747520247.0, "step": 19588 }, { "epoch": 2.4919221473095026, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.083343505859375, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8608487844467163, "num_tokens": 747559983.0, "step": 19589 }, { "epoch": 2.492049357588093, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.062594413757324, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8693790435791016, "num_tokens": 747595746.0, "step": 19590 }, { "epoch": 2.4921765678666836, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.030961036682129, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8714855313301086, "num_tokens": 747639261.0, "step": 19591 }, { "epoch": 2.492303778145274, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.11119556427002, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.855420708656311, "num_tokens": 747680665.0, "step": 19592 }, { "epoch": 2.4924309884238647, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.044634819030762, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8780072927474976, "num_tokens": 747713320.0, "step": 19593 }, { "epoch": 2.492558198702455, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.097647666931152, "learning_rate": 1e-06, "loss": 0.4918, "mean_token_accuracy": 0.8471873998641968, "num_tokens": 747752167.0, "step": 19594 }, { "epoch": 2.4926854089810457, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.05768871307373, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8791916370391846, "num_tokens": 747787274.0, "step": 19595 }, { "epoch": 2.4928126192596363, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.0189790725708, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8679196834564209, "num_tokens": 747826647.0, "step": 19596 }, { "epoch": 2.492939829538227, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.014509201049805, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.869446873664856, "num_tokens": 747866797.0, "step": 19597 }, { "epoch": 2.4930670398168173, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.079787254333496, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8756242990493774, "num_tokens": 747907589.0, "step": 19598 }, { "epoch": 2.493194250095408, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.082074165344238, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8659605979919434, "num_tokens": 747939865.0, "step": 19599 }, { "epoch": 2.4933214603739984, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.06031322479248, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8708434700965881, "num_tokens": 747976238.0, "step": 19600 }, { "epoch": 2.493448670652589, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.056163787841797, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.871224045753479, "num_tokens": 748012076.0, "step": 19601 }, { "epoch": 2.4935758809311794, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.153430938720703, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8668869733810425, "num_tokens": 748042518.0, "step": 19602 }, { "epoch": 2.4937030912097695, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.054681777954102, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8806126117706299, "num_tokens": 748077509.0, "step": 19603 }, { "epoch": 2.4938303014883605, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.04353141784668, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8784655928611755, "num_tokens": 748118164.0, "step": 19604 }, { "epoch": 2.4939575117669506, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.098307609558105, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8675441741943359, "num_tokens": 748153609.0, "step": 19605 }, { "epoch": 2.494084722045541, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.086185455322266, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8711584210395813, "num_tokens": 748189573.0, "step": 19606 }, { "epoch": 2.4942119323241316, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.044154167175293, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8688067197799683, "num_tokens": 748231121.0, "step": 19607 }, { "epoch": 2.494339142602722, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.08915901184082, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8848022818565369, "num_tokens": 748268744.0, "step": 19608 }, { "epoch": 2.4944663528813127, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.057466506958008, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8778177499771118, "num_tokens": 748307441.0, "step": 19609 }, { "epoch": 2.494593563159903, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.035614013671875, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8726838827133179, "num_tokens": 748343506.0, "step": 19610 }, { "epoch": 2.4947207734384937, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.033246994018555, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8740406036376953, "num_tokens": 748382209.0, "step": 19611 }, { "epoch": 2.4948479837170843, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.030406951904297, "learning_rate": 1e-06, "loss": 0.3335, "mean_token_accuracy": 0.8930341005325317, "num_tokens": 748413660.0, "step": 19612 }, { "epoch": 2.494975193995675, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.061168670654297, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8738548755645752, "num_tokens": 748457037.0, "step": 19613 }, { "epoch": 2.4951024042742653, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.035017967224121, "learning_rate": 1e-06, "loss": 0.469, "mean_token_accuracy": 0.8523538112640381, "num_tokens": 748497227.0, "step": 19614 }, { "epoch": 2.495229614552856, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.089557647705078, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8714205026626587, "num_tokens": 748529901.0, "step": 19615 }, { "epoch": 2.4953568248314464, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.07382869720459, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.860312283039093, "num_tokens": 748568238.0, "step": 19616 }, { "epoch": 2.495484035110037, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.0531644821167, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.8635947704315186, "num_tokens": 748608153.0, "step": 19617 }, { "epoch": 2.4956112453886274, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.082887649536133, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8691357374191284, "num_tokens": 748646977.0, "step": 19618 }, { "epoch": 2.495738455667218, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.11521053314209, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8633310794830322, "num_tokens": 748689586.0, "step": 19619 }, { "epoch": 2.4958656659458085, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.059745788574219, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8727207779884338, "num_tokens": 748730915.0, "step": 19620 }, { "epoch": 2.495992876224399, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.090461730957031, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8689547181129456, "num_tokens": 748766996.0, "step": 19621 }, { "epoch": 2.4961200865029896, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.121277809143066, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8654358983039856, "num_tokens": 748801225.0, "step": 19622 }, { "epoch": 2.49624729678158, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.081628799438477, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8680007457733154, "num_tokens": 748841355.0, "step": 19623 }, { "epoch": 2.4963745070601706, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.06740951538086, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.871018648147583, "num_tokens": 748879636.0, "step": 19624 }, { "epoch": 2.496501717338761, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.043402671813965, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8695675134658813, "num_tokens": 748917795.0, "step": 19625 }, { "epoch": 2.4966289276173517, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.090356826782227, "learning_rate": 1e-06, "loss": 0.489, "mean_token_accuracy": 0.8468613624572754, "num_tokens": 748956512.0, "step": 19626 }, { "epoch": 2.496756137895942, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.04373550415039, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.877773106098175, "num_tokens": 748995577.0, "step": 19627 }, { "epoch": 2.4968833481745323, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.124391555786133, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8685780763626099, "num_tokens": 749027672.0, "step": 19628 }, { "epoch": 2.4970105584531233, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.002370834350586, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8835169076919556, "num_tokens": 749059101.0, "step": 19629 }, { "epoch": 2.4971377687317133, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.093184471130371, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8585550785064697, "num_tokens": 749097034.0, "step": 19630 }, { "epoch": 2.497264979010304, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.067856788635254, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.869672417640686, "num_tokens": 749137536.0, "step": 19631 }, { "epoch": 2.4973921892888944, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 13.988916397094727, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8731423616409302, "num_tokens": 749171487.0, "step": 19632 }, { "epoch": 2.497519399567485, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.053237915039062, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8645675182342529, "num_tokens": 749213234.0, "step": 19633 }, { "epoch": 2.4976466098460754, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.055191993713379, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8610877990722656, "num_tokens": 749256184.0, "step": 19634 }, { "epoch": 2.497773820124666, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.068084716796875, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8705800771713257, "num_tokens": 749292909.0, "step": 19635 }, { "epoch": 2.4979010304032565, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.088802337646484, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8596829175949097, "num_tokens": 749327947.0, "step": 19636 }, { "epoch": 2.498028240681847, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.090030670166016, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8811429738998413, "num_tokens": 749364107.0, "step": 19637 }, { "epoch": 2.4981554509604376, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.094051361083984, "learning_rate": 1e-06, "loss": 0.3529, "mean_token_accuracy": 0.887311577796936, "num_tokens": 749398055.0, "step": 19638 }, { "epoch": 2.498282661239028, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.085139274597168, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8727647066116333, "num_tokens": 749434626.0, "step": 19639 }, { "epoch": 2.4984098715176186, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.084670066833496, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8656418919563293, "num_tokens": 749470616.0, "step": 19640 }, { "epoch": 2.498537081796209, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.037981986999512, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8749439716339111, "num_tokens": 749513072.0, "step": 19641 }, { "epoch": 2.4986642920747997, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.149177551269531, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8864408731460571, "num_tokens": 749546988.0, "step": 19642 }, { "epoch": 2.49879150235339, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.0923433303833, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.8665039539337158, "num_tokens": 749585851.0, "step": 19643 }, { "epoch": 2.4989187126319807, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.058664321899414, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8710164427757263, "num_tokens": 749619510.0, "step": 19644 }, { "epoch": 2.4990459229105713, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.093027114868164, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.880527913570404, "num_tokens": 749653660.0, "step": 19645 }, { "epoch": 2.499173133189162, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.046708106994629, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8727583885192871, "num_tokens": 749691321.0, "step": 19646 }, { "epoch": 2.4993003434677523, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.027132034301758, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8737190365791321, "num_tokens": 749729305.0, "step": 19647 }, { "epoch": 2.499427553746343, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.088569641113281, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8782029151916504, "num_tokens": 749766134.0, "step": 19648 }, { "epoch": 2.4995547640249334, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.0682373046875, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8696194291114807, "num_tokens": 749807715.0, "step": 19649 }, { "epoch": 2.499681974303524, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.150957107543945, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8744416832923889, "num_tokens": 749838799.0, "step": 19650 }, { "epoch": 2.499809184582114, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.022268295288086, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8857435584068298, "num_tokens": 749874283.0, "step": 19651 }, { "epoch": 2.499936394860705, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.098624229431152, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8705278635025024, "num_tokens": 749915912.0, "step": 19652 }, { "epoch": 2.500063605139295, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.03173542022705, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8789818286895752, "num_tokens": 749959133.0, "step": 19653 }, { "epoch": 2.500190815417886, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.060652732849121, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8697174787521362, "num_tokens": 749997846.0, "step": 19654 }, { "epoch": 2.500318025696476, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.072025299072266, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8763279318809509, "num_tokens": 750040973.0, "step": 19655 }, { "epoch": 2.5004452359750666, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.064337730407715, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8670928478240967, "num_tokens": 750082171.0, "step": 19656 }, { "epoch": 2.500572446253657, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.14050006866455, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8839393854141235, "num_tokens": 750123083.0, "step": 19657 }, { "epoch": 2.5006996565322477, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.025260925292969, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8583971261978149, "num_tokens": 750162235.0, "step": 19658 }, { "epoch": 2.500826866810838, "ewc_loss": 0.037353515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 14.068315505981445, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8845947980880737, "num_tokens": 750196501.0, "step": 19659 }, { "epoch": 2.5009540770894287, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.0860595703125, "learning_rate": 1e-06, "loss": 0.5008, "mean_token_accuracy": 0.8449398279190063, "num_tokens": 750241178.0, "step": 19660 }, { "epoch": 2.5010812873680193, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.052452087402344, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8776859045028687, "num_tokens": 750284104.0, "step": 19661 }, { "epoch": 2.50120849764661, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.102195739746094, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8781632781028748, "num_tokens": 750318525.0, "step": 19662 }, { "epoch": 2.5013357079252003, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.07618236541748, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8813016414642334, "num_tokens": 750358892.0, "step": 19663 }, { "epoch": 2.501462918203791, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.098315238952637, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8602291345596313, "num_tokens": 750399325.0, "step": 19664 }, { "epoch": 2.5015901284823814, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.083118438720703, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8658981919288635, "num_tokens": 750437493.0, "step": 19665 }, { "epoch": 2.501717338760972, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.095276832580566, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.880638837814331, "num_tokens": 750472381.0, "step": 19666 }, { "epoch": 2.5018445490395624, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.086485862731934, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.856471836566925, "num_tokens": 750511786.0, "step": 19667 }, { "epoch": 2.501971759318153, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.08256721496582, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8793216943740845, "num_tokens": 750550596.0, "step": 19668 }, { "epoch": 2.5020989695967435, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.01420783996582, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8767597675323486, "num_tokens": 750594790.0, "step": 19669 }, { "epoch": 2.502226179875334, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.108453750610352, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8850008249282837, "num_tokens": 750636738.0, "step": 19670 }, { "epoch": 2.5023533901539246, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.03428840637207, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8849188089370728, "num_tokens": 750672338.0, "step": 19671 }, { "epoch": 2.502480600432515, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.02261734008789, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8668437004089355, "num_tokens": 750709149.0, "step": 19672 }, { "epoch": 2.5026078107111056, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.112783432006836, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.885403037071228, "num_tokens": 750741727.0, "step": 19673 }, { "epoch": 2.5027350209896957, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.052379608154297, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8601423501968384, "num_tokens": 750775903.0, "step": 19674 }, { "epoch": 2.5028622312682867, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.067313194274902, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8685881495475769, "num_tokens": 750810502.0, "step": 19675 }, { "epoch": 2.5029894415468767, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.081865310668945, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8803879022598267, "num_tokens": 750847762.0, "step": 19676 }, { "epoch": 2.5031166518254677, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.028785705566406, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8741642832756042, "num_tokens": 750887201.0, "step": 19677 }, { "epoch": 2.503243862104058, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.13162899017334, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8712139129638672, "num_tokens": 750927323.0, "step": 19678 }, { "epoch": 2.5033710723826488, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.079207420349121, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8644429445266724, "num_tokens": 750972501.0, "step": 19679 }, { "epoch": 2.503498282661239, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.062487602233887, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8826569318771362, "num_tokens": 751010743.0, "step": 19680 }, { "epoch": 2.5036254929398294, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.117101669311523, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8685654997825623, "num_tokens": 751050440.0, "step": 19681 }, { "epoch": 2.50375270321842, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.10982608795166, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8766562342643738, "num_tokens": 751088319.0, "step": 19682 }, { "epoch": 2.5038799134970104, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.09738826751709, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8729131817817688, "num_tokens": 751123766.0, "step": 19683 }, { "epoch": 2.504007123775601, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.068765640258789, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8883293867111206, "num_tokens": 751163375.0, "step": 19684 }, { "epoch": 2.5041343340541915, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.107513427734375, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8680055141448975, "num_tokens": 751204526.0, "step": 19685 }, { "epoch": 2.504261544332782, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.073343276977539, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8783077001571655, "num_tokens": 751245133.0, "step": 19686 }, { "epoch": 2.5043887546113726, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.132494926452637, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8628937005996704, "num_tokens": 751284578.0, "step": 19687 }, { "epoch": 2.504515964889963, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.039114952087402, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8640780448913574, "num_tokens": 751325794.0, "step": 19688 }, { "epoch": 2.5046431751685536, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.060160636901855, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8794960379600525, "num_tokens": 751363777.0, "step": 19689 }, { "epoch": 2.504770385447144, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.057978630065918, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8633360862731934, "num_tokens": 751400499.0, "step": 19690 }, { "epoch": 2.5048975957257347, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.076414108276367, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8709689974784851, "num_tokens": 751434895.0, "step": 19691 }, { "epoch": 2.505024806004325, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.025520324707031, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8823585510253906, "num_tokens": 751473984.0, "step": 19692 }, { "epoch": 2.5051520162829157, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.064589500427246, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8738176226615906, "num_tokens": 751515030.0, "step": 19693 }, { "epoch": 2.5052792265615063, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.120206832885742, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8705762624740601, "num_tokens": 751550023.0, "step": 19694 }, { "epoch": 2.505406436840097, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.070408821105957, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.874519944190979, "num_tokens": 751584793.0, "step": 19695 }, { "epoch": 2.5055336471186873, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.073075294494629, "learning_rate": 1e-06, "loss": 0.4546, "mean_token_accuracy": 0.8548442125320435, "num_tokens": 751619817.0, "step": 19696 }, { "epoch": 2.505660857397278, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.008415222167969, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8786954879760742, "num_tokens": 751657097.0, "step": 19697 }, { "epoch": 2.5057880676758684, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 14.110225677490234, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8641250133514404, "num_tokens": 751687042.0, "step": 19698 }, { "epoch": 2.5059152779544585, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.006386756896973, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8804479837417603, "num_tokens": 751724281.0, "step": 19699 }, { "epoch": 2.5060424882330494, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.045856475830078, "learning_rate": 1e-06, "loss": 0.456, "mean_token_accuracy": 0.8536344766616821, "num_tokens": 751765144.0, "step": 19700 }, { "epoch": 2.5061696985116395, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.005456924438477, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8798284530639648, "num_tokens": 751799602.0, "step": 19701 }, { "epoch": 2.5062969087902305, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.130188941955566, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8817045092582703, "num_tokens": 751836972.0, "step": 19702 }, { "epoch": 2.5064241190688206, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.016165733337402, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8738741874694824, "num_tokens": 751869627.0, "step": 19703 }, { "epoch": 2.5065513293474115, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.084758758544922, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8653278350830078, "num_tokens": 751911103.0, "step": 19704 }, { "epoch": 2.5066785396260016, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.021086692810059, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8676186800003052, "num_tokens": 751948956.0, "step": 19705 }, { "epoch": 2.506805749904592, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.027230262756348, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8761526346206665, "num_tokens": 751992185.0, "step": 19706 }, { "epoch": 2.5069329601831827, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.000967979431152, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8695434927940369, "num_tokens": 752032661.0, "step": 19707 }, { "epoch": 2.507060170461773, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.050917625427246, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8779325485229492, "num_tokens": 752073946.0, "step": 19708 }, { "epoch": 2.5071873807403637, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.042089462280273, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8586363792419434, "num_tokens": 752118766.0, "step": 19709 }, { "epoch": 2.5073145910189543, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.080751419067383, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8758795261383057, "num_tokens": 752159669.0, "step": 19710 }, { "epoch": 2.507441801297545, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.046648025512695, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8769252896308899, "num_tokens": 752198139.0, "step": 19711 }, { "epoch": 2.5075690115761353, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.086409568786621, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8753386735916138, "num_tokens": 752239686.0, "step": 19712 }, { "epoch": 2.507696221854726, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.089103698730469, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8549245595932007, "num_tokens": 752275740.0, "step": 19713 }, { "epoch": 2.5078234321333164, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.111772537231445, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.861027717590332, "num_tokens": 752310554.0, "step": 19714 }, { "epoch": 2.507950642411907, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.020756721496582, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8774863481521606, "num_tokens": 752347761.0, "step": 19715 }, { "epoch": 2.5080778526904974, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.11546802520752, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8630155920982361, "num_tokens": 752387128.0, "step": 19716 }, { "epoch": 2.508205062969088, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.08474349975586, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8719930648803711, "num_tokens": 752428722.0, "step": 19717 }, { "epoch": 2.5083322732476785, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.06159782409668, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8714597225189209, "num_tokens": 752468920.0, "step": 19718 }, { "epoch": 2.508459483526269, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.120949745178223, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.863275945186615, "num_tokens": 752505226.0, "step": 19719 }, { "epoch": 2.5085866938048595, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.083550453186035, "learning_rate": 1e-06, "loss": 0.3927, "mean_token_accuracy": 0.873729944229126, "num_tokens": 752543693.0, "step": 19720 }, { "epoch": 2.50871390408345, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.103325843811035, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8588045835494995, "num_tokens": 752578758.0, "step": 19721 }, { "epoch": 2.5088411143620406, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.06843376159668, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8828485608100891, "num_tokens": 752612115.0, "step": 19722 }, { "epoch": 2.508968324640631, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.066842079162598, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8778356909751892, "num_tokens": 752644601.0, "step": 19723 }, { "epoch": 2.509095534919221, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.005666732788086, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8750423192977905, "num_tokens": 752682333.0, "step": 19724 }, { "epoch": 2.509222745197812, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.073287963867188, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8670544624328613, "num_tokens": 752721802.0, "step": 19725 }, { "epoch": 2.5093499554764023, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.087242126464844, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8874316811561584, "num_tokens": 752755673.0, "step": 19726 }, { "epoch": 2.5094771657549932, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.015108108520508, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8821445107460022, "num_tokens": 752795575.0, "step": 19727 }, { "epoch": 2.5096043760335833, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.105172157287598, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8886067271232605, "num_tokens": 752833172.0, "step": 19728 }, { "epoch": 2.5097315863121743, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.08524227142334, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8580524325370789, "num_tokens": 752872857.0, "step": 19729 }, { "epoch": 2.5098587965907644, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.153865814208984, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8823597431182861, "num_tokens": 752908553.0, "step": 19730 }, { "epoch": 2.509986006869355, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.036505699157715, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.877468466758728, "num_tokens": 752937762.0, "step": 19731 }, { "epoch": 2.5101132171479454, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.113547325134277, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8591877222061157, "num_tokens": 752977184.0, "step": 19732 }, { "epoch": 2.510240427426536, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.063691139221191, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8687228560447693, "num_tokens": 753021071.0, "step": 19733 }, { "epoch": 2.5103676377051265, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.082242965698242, "learning_rate": 1e-06, "loss": 0.4843, "mean_token_accuracy": 0.8473312258720398, "num_tokens": 753056597.0, "step": 19734 }, { "epoch": 2.510494847983717, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.105534553527832, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8758101463317871, "num_tokens": 753096178.0, "step": 19735 }, { "epoch": 2.5106220582623076, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.042667388916016, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8760582208633423, "num_tokens": 753130388.0, "step": 19736 }, { "epoch": 2.510749268540898, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.054454803466797, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8806124925613403, "num_tokens": 753166585.0, "step": 19737 }, { "epoch": 2.5108764788194886, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.03315258026123, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8815876245498657, "num_tokens": 753203403.0, "step": 19738 }, { "epoch": 2.511003689098079, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.029926300048828, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8784099817276001, "num_tokens": 753241632.0, "step": 19739 }, { "epoch": 2.5111308993766697, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.041890144348145, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.88620924949646, "num_tokens": 753279292.0, "step": 19740 }, { "epoch": 2.51125810965526, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.101316452026367, "learning_rate": 1e-06, "loss": 0.4298, "mean_token_accuracy": 0.8639913201332092, "num_tokens": 753319121.0, "step": 19741 }, { "epoch": 2.5113853199338507, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.103498458862305, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8576231002807617, "num_tokens": 753353074.0, "step": 19742 }, { "epoch": 2.5115125302124413, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.025654792785645, "learning_rate": 1e-06, "loss": 0.3475, "mean_token_accuracy": 0.8907637596130371, "num_tokens": 753388994.0, "step": 19743 }, { "epoch": 2.511639740491032, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.141521453857422, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8792337775230408, "num_tokens": 753430271.0, "step": 19744 }, { "epoch": 2.5117669507696223, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.044238090515137, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8657760620117188, "num_tokens": 753473820.0, "step": 19745 }, { "epoch": 2.511894161048213, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.125184059143066, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8691447973251343, "num_tokens": 753516397.0, "step": 19746 }, { "epoch": 2.5120213713268034, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.09444808959961, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.882306694984436, "num_tokens": 753557589.0, "step": 19747 }, { "epoch": 2.512148581605394, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.101081848144531, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8806247711181641, "num_tokens": 753593764.0, "step": 19748 }, { "epoch": 2.512275791883984, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.089441299438477, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8605168461799622, "num_tokens": 753635967.0, "step": 19749 }, { "epoch": 2.512403002162575, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.043906211853027, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8812049031257629, "num_tokens": 753674595.0, "step": 19750 }, { "epoch": 2.512530212441165, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.073295593261719, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.85699063539505, "num_tokens": 753714118.0, "step": 19751 }, { "epoch": 2.512657422719756, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.105140686035156, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8587406277656555, "num_tokens": 753758139.0, "step": 19752 }, { "epoch": 2.512784632998346, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.111200332641602, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8615890741348267, "num_tokens": 753800373.0, "step": 19753 }, { "epoch": 2.5129118432769366, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.114313125610352, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8669209480285645, "num_tokens": 753832366.0, "step": 19754 }, { "epoch": 2.513039053555527, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.044811248779297, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.890839695930481, "num_tokens": 753866176.0, "step": 19755 }, { "epoch": 2.5131662638341177, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.116957664489746, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8716341257095337, "num_tokens": 753911470.0, "step": 19756 }, { "epoch": 2.513293474112708, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.090462684631348, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8702433705329895, "num_tokens": 753953027.0, "step": 19757 }, { "epoch": 2.5134206843912987, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.153728485107422, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8664200305938721, "num_tokens": 753987992.0, "step": 19758 }, { "epoch": 2.5135478946698893, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.1436128616333, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8770956993103027, "num_tokens": 754026671.0, "step": 19759 }, { "epoch": 2.51367510494848, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.06531047821045, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8735446929931641, "num_tokens": 754063526.0, "step": 19760 }, { "epoch": 2.5138023152270703, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.106505393981934, "learning_rate": 1e-06, "loss": 0.4505, "mean_token_accuracy": 0.8553028702735901, "num_tokens": 754103361.0, "step": 19761 }, { "epoch": 2.513929525505661, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.094159126281738, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8755865097045898, "num_tokens": 754141969.0, "step": 19762 }, { "epoch": 2.5140567357842514, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.036441802978516, "learning_rate": 1e-06, "loss": 0.3541, "mean_token_accuracy": 0.8886308670043945, "num_tokens": 754182691.0, "step": 19763 }, { "epoch": 2.514183946062842, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.121206283569336, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8719441890716553, "num_tokens": 754219956.0, "step": 19764 }, { "epoch": 2.5143111563414324, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.050148010253906, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8686689138412476, "num_tokens": 754259467.0, "step": 19765 }, { "epoch": 2.514438366620023, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.100143432617188, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.876091718673706, "num_tokens": 754300344.0, "step": 19766 }, { "epoch": 2.5145655768986135, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.040975570678711, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8719005584716797, "num_tokens": 754342606.0, "step": 19767 }, { "epoch": 2.514692787177204, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.13025951385498, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8682845830917358, "num_tokens": 754379946.0, "step": 19768 }, { "epoch": 2.5148199974557945, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.021062850952148, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8865771889686584, "num_tokens": 754414353.0, "step": 19769 }, { "epoch": 2.514947207734385, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.102201461791992, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8603715896606445, "num_tokens": 754455506.0, "step": 19770 }, { "epoch": 2.5150744180129756, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.046451568603516, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8563299179077148, "num_tokens": 754498316.0, "step": 19771 }, { "epoch": 2.5152016282915657, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.078417778015137, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8612744808197021, "num_tokens": 754534943.0, "step": 19772 }, { "epoch": 2.5153288385701567, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.09426212310791, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8704224824905396, "num_tokens": 754572775.0, "step": 19773 }, { "epoch": 2.5154560488487467, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.038430213928223, "learning_rate": 1e-06, "loss": 0.3583, "mean_token_accuracy": 0.882442831993103, "num_tokens": 754610473.0, "step": 19774 }, { "epoch": 2.5155832591273377, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.094465255737305, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8734414577484131, "num_tokens": 754649527.0, "step": 19775 }, { "epoch": 2.515710469405928, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.078822135925293, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8746976256370544, "num_tokens": 754691513.0, "step": 19776 }, { "epoch": 2.5158376796845188, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.03781509399414, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8602354526519775, "num_tokens": 754727932.0, "step": 19777 }, { "epoch": 2.515964889963109, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.056525230407715, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8782246708869934, "num_tokens": 754764571.0, "step": 19778 }, { "epoch": 2.5160921002416994, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.0717134475708, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8868359327316284, "num_tokens": 754805846.0, "step": 19779 }, { "epoch": 2.51621931052029, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.077352523803711, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8806317448616028, "num_tokens": 754844016.0, "step": 19780 }, { "epoch": 2.5163465207988804, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.097145080566406, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8723308444023132, "num_tokens": 754886493.0, "step": 19781 }, { "epoch": 2.516473731077471, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.124897003173828, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8750625848770142, "num_tokens": 754928195.0, "step": 19782 }, { "epoch": 2.5166009413560615, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.0301513671875, "learning_rate": 1e-06, "loss": 0.352, "mean_token_accuracy": 0.8818379640579224, "num_tokens": 754958399.0, "step": 19783 }, { "epoch": 2.516728151634652, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.086684226989746, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8769767880439758, "num_tokens": 754997926.0, "step": 19784 }, { "epoch": 2.5168553619132426, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.006717681884766, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8784026503562927, "num_tokens": 755040577.0, "step": 19785 }, { "epoch": 2.516982572191833, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.135774612426758, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.87059485912323, "num_tokens": 755070458.0, "step": 19786 }, { "epoch": 2.5171097824704236, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.059321403503418, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8521018624305725, "num_tokens": 755117205.0, "step": 19787 }, { "epoch": 2.517236992749014, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.103425979614258, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8743065595626831, "num_tokens": 755162098.0, "step": 19788 }, { "epoch": 2.5173642030276047, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.096426963806152, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8721266984939575, "num_tokens": 755199655.0, "step": 19789 }, { "epoch": 2.517491413306195, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.073987007141113, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8861465454101562, "num_tokens": 755234165.0, "step": 19790 }, { "epoch": 2.5176186235847857, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.105266571044922, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8744751214981079, "num_tokens": 755271933.0, "step": 19791 }, { "epoch": 2.5177458338633762, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.07712173461914, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8823443651199341, "num_tokens": 755306903.0, "step": 19792 }, { "epoch": 2.5178730441419668, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.099129676818848, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8809440732002258, "num_tokens": 755347613.0, "step": 19793 }, { "epoch": 2.5180002544205573, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.046807289123535, "learning_rate": 1e-06, "loss": 0.3471, "mean_token_accuracy": 0.8912405967712402, "num_tokens": 755380541.0, "step": 19794 }, { "epoch": 2.518127464699148, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.096844673156738, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8706167936325073, "num_tokens": 755422502.0, "step": 19795 }, { "epoch": 2.5182546749777384, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.032214164733887, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8749046921730042, "num_tokens": 755463975.0, "step": 19796 }, { "epoch": 2.5183818852563284, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.127514839172363, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8746675848960876, "num_tokens": 755504873.0, "step": 19797 }, { "epoch": 2.5185090955349194, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.096620559692383, "learning_rate": 1e-06, "loss": 0.4784, "mean_token_accuracy": 0.8489564657211304, "num_tokens": 755545646.0, "step": 19798 }, { "epoch": 2.5186363058135095, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.150579452514648, "learning_rate": 1e-06, "loss": 0.4431, "mean_token_accuracy": 0.8600490689277649, "num_tokens": 755577228.0, "step": 19799 }, { "epoch": 2.5187635160921005, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.084348678588867, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8811150193214417, "num_tokens": 755621437.0, "step": 19800 }, { "epoch": 2.5188907263706906, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.077396392822266, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8706613779067993, "num_tokens": 755658266.0, "step": 19801 }, { "epoch": 2.5190179366492815, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.073261260986328, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8618953227996826, "num_tokens": 755696755.0, "step": 19802 }, { "epoch": 2.5191451469278716, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.037487983703613, "learning_rate": 1e-06, "loss": 0.3531, "mean_token_accuracy": 0.8862755298614502, "num_tokens": 755732434.0, "step": 19803 }, { "epoch": 2.519272357206462, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.104726791381836, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8780577778816223, "num_tokens": 755768488.0, "step": 19804 }, { "epoch": 2.5193995674850527, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.103048324584961, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8797006011009216, "num_tokens": 755799637.0, "step": 19805 }, { "epoch": 2.519526777763643, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.084117889404297, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.857588529586792, "num_tokens": 755834229.0, "step": 19806 }, { "epoch": 2.5196539880422337, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.123138427734375, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8767685294151306, "num_tokens": 755871776.0, "step": 19807 }, { "epoch": 2.5197811983208243, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.071765899658203, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8838695883750916, "num_tokens": 755911125.0, "step": 19808 }, { "epoch": 2.519908408599415, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.08272933959961, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8817201852798462, "num_tokens": 755952204.0, "step": 19809 }, { "epoch": 2.5200356188780053, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.066313743591309, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8627932071685791, "num_tokens": 755996702.0, "step": 19810 }, { "epoch": 2.520162829156596, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.084596633911133, "learning_rate": 1e-06, "loss": 0.3533, "mean_token_accuracy": 0.8897922039031982, "num_tokens": 756033285.0, "step": 19811 }, { "epoch": 2.5202900394351864, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.098691940307617, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8728132843971252, "num_tokens": 756068851.0, "step": 19812 }, { "epoch": 2.520417249713777, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.140433311462402, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.86501145362854, "num_tokens": 756111676.0, "step": 19813 }, { "epoch": 2.5205444599923674, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.093799591064453, "learning_rate": 1e-06, "loss": 0.4577, "mean_token_accuracy": 0.8578535318374634, "num_tokens": 756146575.0, "step": 19814 }, { "epoch": 2.520671670270958, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.08545207977295, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8629857301712036, "num_tokens": 756190006.0, "step": 19815 }, { "epoch": 2.5207988805495485, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.042030334472656, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8838194608688354, "num_tokens": 756232828.0, "step": 19816 }, { "epoch": 2.520926090828139, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.138915061950684, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8797692656517029, "num_tokens": 756273077.0, "step": 19817 }, { "epoch": 2.5210533011067295, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.12270736694336, "learning_rate": 1e-06, "loss": 0.4551, "mean_token_accuracy": 0.8578764200210571, "num_tokens": 756313421.0, "step": 19818 }, { "epoch": 2.52118051138532, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.089110374450684, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8691054582595825, "num_tokens": 756356461.0, "step": 19819 }, { "epoch": 2.5213077216639106, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.191914558410645, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8696807622909546, "num_tokens": 756388309.0, "step": 19820 }, { "epoch": 2.521434931942501, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.087174415588379, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.883934497833252, "num_tokens": 756426655.0, "step": 19821 }, { "epoch": 2.521562142221091, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.101381301879883, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.870023787021637, "num_tokens": 756464961.0, "step": 19822 }, { "epoch": 2.521689352499682, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.145841598510742, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8884400129318237, "num_tokens": 756497338.0, "step": 19823 }, { "epoch": 2.5218165627782723, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.143583297729492, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8791245818138123, "num_tokens": 756535324.0, "step": 19824 }, { "epoch": 2.5219437730568632, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.078699111938477, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8810737133026123, "num_tokens": 756574138.0, "step": 19825 }, { "epoch": 2.5220709833354533, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.1465482711792, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8826350569725037, "num_tokens": 756608967.0, "step": 19826 }, { "epoch": 2.522198193614044, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.061132431030273, "learning_rate": 1e-06, "loss": 0.4727, "mean_token_accuracy": 0.8532012701034546, "num_tokens": 756653882.0, "step": 19827 }, { "epoch": 2.5223254038926344, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.081823348999023, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8685595989227295, "num_tokens": 756695638.0, "step": 19828 }, { "epoch": 2.522452614171225, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.131936073303223, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8663231134414673, "num_tokens": 756734964.0, "step": 19829 }, { "epoch": 2.5225798244498154, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.094573020935059, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8775279521942139, "num_tokens": 756768454.0, "step": 19830 }, { "epoch": 2.522707034728406, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.106773376464844, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8622337579727173, "num_tokens": 756801696.0, "step": 19831 }, { "epoch": 2.5228342450069965, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.146459579467773, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8737711310386658, "num_tokens": 756839874.0, "step": 19832 }, { "epoch": 2.522961455285587, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.025657653808594, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.862449049949646, "num_tokens": 756884179.0, "step": 19833 }, { "epoch": 2.5230886655641775, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.139835357666016, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8748385906219482, "num_tokens": 756922725.0, "step": 19834 }, { "epoch": 2.523215875842768, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.132621765136719, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8684961795806885, "num_tokens": 756956080.0, "step": 19835 }, { "epoch": 2.5233430861213586, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.099103927612305, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8637539744377136, "num_tokens": 756995137.0, "step": 19836 }, { "epoch": 2.523470296399949, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.103272438049316, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8792405128479004, "num_tokens": 757035300.0, "step": 19837 }, { "epoch": 2.5235975066785397, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.130180358886719, "learning_rate": 1e-06, "loss": 0.4772, "mean_token_accuracy": 0.8493648171424866, "num_tokens": 757068789.0, "step": 19838 }, { "epoch": 2.52372471695713, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.105122566223145, "learning_rate": 1e-06, "loss": 0.4318, "mean_token_accuracy": 0.8679251074790955, "num_tokens": 757102913.0, "step": 19839 }, { "epoch": 2.5238519272357207, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.08565902709961, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8745790719985962, "num_tokens": 757138932.0, "step": 19840 }, { "epoch": 2.5239791375143112, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.07298755645752, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8757643699645996, "num_tokens": 757179181.0, "step": 19841 }, { "epoch": 2.5241063477929018, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.092844009399414, "learning_rate": 1e-06, "loss": 0.4902, "mean_token_accuracy": 0.8428511619567871, "num_tokens": 757224209.0, "step": 19842 }, { "epoch": 2.5242335580714923, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.082049369812012, "learning_rate": 1e-06, "loss": 0.3574, "mean_token_accuracy": 0.8854261636734009, "num_tokens": 757262398.0, "step": 19843 }, { "epoch": 2.524360768350083, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.074963569641113, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8632493019104004, "num_tokens": 757301744.0, "step": 19844 }, { "epoch": 2.5244879786286734, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.10170841217041, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8774960041046143, "num_tokens": 757338273.0, "step": 19845 }, { "epoch": 2.524615188907264, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.097177505493164, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.86538165807724, "num_tokens": 757383169.0, "step": 19846 }, { "epoch": 2.524742399185854, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.014679908752441, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8621867895126343, "num_tokens": 757420998.0, "step": 19847 }, { "epoch": 2.524869609464445, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.117705345153809, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8771904706954956, "num_tokens": 757463281.0, "step": 19848 }, { "epoch": 2.524996819743035, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.06598949432373, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8676484227180481, "num_tokens": 757495022.0, "step": 19849 }, { "epoch": 2.525124030021626, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.1223726272583, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8783149719238281, "num_tokens": 757536801.0, "step": 19850 }, { "epoch": 2.525251240300216, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.048213005065918, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8871932029724121, "num_tokens": 757573419.0, "step": 19851 }, { "epoch": 2.5253784505788066, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.113429069519043, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8671190142631531, "num_tokens": 757610773.0, "step": 19852 }, { "epoch": 2.525505660857397, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.058647155761719, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8825361132621765, "num_tokens": 757648067.0, "step": 19853 }, { "epoch": 2.5256328711359877, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.115377426147461, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8790068626403809, "num_tokens": 757681598.0, "step": 19854 }, { "epoch": 2.525760081414578, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.123292922973633, "learning_rate": 1e-06, "loss": 0.4754, "mean_token_accuracy": 0.8547871708869934, "num_tokens": 757724752.0, "step": 19855 }, { "epoch": 2.5258872916931687, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.026583671569824, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.8848052024841309, "num_tokens": 757763801.0, "step": 19856 }, { "epoch": 2.5260145019717593, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.126663208007812, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8712139129638672, "num_tokens": 757799235.0, "step": 19857 }, { "epoch": 2.52614171225035, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.123448371887207, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8770081996917725, "num_tokens": 757839695.0, "step": 19858 }, { "epoch": 2.5262689225289403, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.100545883178711, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8593778610229492, "num_tokens": 757877152.0, "step": 19859 }, { "epoch": 2.526396132807531, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.115364074707031, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8721678256988525, "num_tokens": 757916173.0, "step": 19860 }, { "epoch": 2.5265233430861214, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.111023902893066, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8783214688301086, "num_tokens": 757957271.0, "step": 19861 }, { "epoch": 2.526650553364712, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.146171569824219, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8671020269393921, "num_tokens": 757995553.0, "step": 19862 }, { "epoch": 2.5267777636433024, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.06719970703125, "learning_rate": 1e-06, "loss": 0.4413, "mean_token_accuracy": 0.8617189526557922, "num_tokens": 758035095.0, "step": 19863 }, { "epoch": 2.526904973921893, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.165809631347656, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8706244230270386, "num_tokens": 758071185.0, "step": 19864 }, { "epoch": 2.5270321842004835, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.08714771270752, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8729032278060913, "num_tokens": 758110251.0, "step": 19865 }, { "epoch": 2.527159394479074, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.175619125366211, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8586469888687134, "num_tokens": 758144385.0, "step": 19866 }, { "epoch": 2.5272866047576645, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.077995300292969, "learning_rate": 1e-06, "loss": 0.3597, "mean_token_accuracy": 0.8839510679244995, "num_tokens": 758185935.0, "step": 19867 }, { "epoch": 2.527413815036255, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.089938163757324, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8765488862991333, "num_tokens": 758226688.0, "step": 19868 }, { "epoch": 2.5275410253148456, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.194327354431152, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8776347637176514, "num_tokens": 758263866.0, "step": 19869 }, { "epoch": 2.5276682355934357, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.069568634033203, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8781158924102783, "num_tokens": 758303607.0, "step": 19870 }, { "epoch": 2.5277954458720266, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 14.10154914855957, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8765435218811035, "num_tokens": 758340120.0, "step": 19871 }, { "epoch": 2.5279226561506167, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.128540992736816, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.872851550579071, "num_tokens": 758373792.0, "step": 19872 }, { "epoch": 2.5280498664292077, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.070393562316895, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8617246150970459, "num_tokens": 758418157.0, "step": 19873 }, { "epoch": 2.528177076707798, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.06459903717041, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8652023077011108, "num_tokens": 758454042.0, "step": 19874 }, { "epoch": 2.5283042869863888, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.0931396484375, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8742688894271851, "num_tokens": 758495305.0, "step": 19875 }, { "epoch": 2.528431497264979, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.076616287231445, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8799883723258972, "num_tokens": 758532646.0, "step": 19876 }, { "epoch": 2.5285587075435694, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.107311248779297, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8683789968490601, "num_tokens": 758568329.0, "step": 19877 }, { "epoch": 2.52868591782216, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.084453582763672, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8753959536552429, "num_tokens": 758608246.0, "step": 19878 }, { "epoch": 2.5288131281007504, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.09453010559082, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.882210910320282, "num_tokens": 758645356.0, "step": 19879 }, { "epoch": 2.528940338379341, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.01504135131836, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8879409432411194, "num_tokens": 758685657.0, "step": 19880 }, { "epoch": 2.5290675486579315, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.10519027709961, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8759604096412659, "num_tokens": 758718843.0, "step": 19881 }, { "epoch": 2.529194758936522, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.179527282714844, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8775693774223328, "num_tokens": 758757946.0, "step": 19882 }, { "epoch": 2.5293219692151125, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.118534088134766, "learning_rate": 1e-06, "loss": 0.4644, "mean_token_accuracy": 0.8544599413871765, "num_tokens": 758792832.0, "step": 19883 }, { "epoch": 2.529449179493703, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.201065063476562, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.865564227104187, "num_tokens": 758826062.0, "step": 19884 }, { "epoch": 2.5295763897722936, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.14655876159668, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8821691274642944, "num_tokens": 758860940.0, "step": 19885 }, { "epoch": 2.529703600050884, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.09289836883545, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.871701717376709, "num_tokens": 758900740.0, "step": 19886 }, { "epoch": 2.5298308103294747, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.14533805847168, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8700392842292786, "num_tokens": 758942786.0, "step": 19887 }, { "epoch": 2.529958020608065, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.083517074584961, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8816112875938416, "num_tokens": 758983376.0, "step": 19888 }, { "epoch": 2.5300852308866557, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.131383895874023, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8618865013122559, "num_tokens": 759024007.0, "step": 19889 }, { "epoch": 2.5302124411652462, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.072672843933105, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8808021545410156, "num_tokens": 759061611.0, "step": 19890 }, { "epoch": 2.5303396514438368, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.169233322143555, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8572858572006226, "num_tokens": 759101942.0, "step": 19891 }, { "epoch": 2.5304668617224273, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.104069709777832, "learning_rate": 1e-06, "loss": 0.3492, "mean_token_accuracy": 0.8903579711914062, "num_tokens": 759140281.0, "step": 19892 }, { "epoch": 2.530594072001018, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.115492820739746, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8727653622627258, "num_tokens": 759174319.0, "step": 19893 }, { "epoch": 2.5307212822796084, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.155221939086914, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8794374465942383, "num_tokens": 759208981.0, "step": 19894 }, { "epoch": 2.5308484925581984, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.129206657409668, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8739675879478455, "num_tokens": 759240559.0, "step": 19895 }, { "epoch": 2.5309757028367894, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.152848243713379, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8849210739135742, "num_tokens": 759280550.0, "step": 19896 }, { "epoch": 2.5311029131153795, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.100207328796387, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8666598796844482, "num_tokens": 759321298.0, "step": 19897 }, { "epoch": 2.5312301233939705, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.175405502319336, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8852453231811523, "num_tokens": 759354627.0, "step": 19898 }, { "epoch": 2.5313573336725606, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.133171081542969, "learning_rate": 1e-06, "loss": 0.4207, "mean_token_accuracy": 0.8675680160522461, "num_tokens": 759388129.0, "step": 19899 }, { "epoch": 2.5314845439511515, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.155570983886719, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8818920254707336, "num_tokens": 759423667.0, "step": 19900 }, { "epoch": 2.5316117542297416, "ewc_loss": 0.03759765625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.074363708496094, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8639742732048035, "num_tokens": 759466586.0, "step": 19901 }, { "epoch": 2.531738964508332, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.101094245910645, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8674275279045105, "num_tokens": 759504214.0, "step": 19902 }, { "epoch": 2.5318661747869227, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.091591835021973, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8598116636276245, "num_tokens": 759543865.0, "step": 19903 }, { "epoch": 2.531993385065513, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.121078491210938, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8679970502853394, "num_tokens": 759581917.0, "step": 19904 }, { "epoch": 2.5321205953441037, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.117300033569336, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8754711151123047, "num_tokens": 759617641.0, "step": 19905 }, { "epoch": 2.5322478056226942, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.140007972717285, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8837332725524902, "num_tokens": 759649097.0, "step": 19906 }, { "epoch": 2.5323750159012848, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.13865852355957, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8687254190444946, "num_tokens": 759688802.0, "step": 19907 }, { "epoch": 2.5325022261798753, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.086928367614746, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8818632364273071, "num_tokens": 759724243.0, "step": 19908 }, { "epoch": 2.532629436458466, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.126673698425293, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8669707179069519, "num_tokens": 759764646.0, "step": 19909 }, { "epoch": 2.5327566467370564, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.079811096191406, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8760101199150085, "num_tokens": 759806939.0, "step": 19910 }, { "epoch": 2.532883857015647, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.134101867675781, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8755166530609131, "num_tokens": 759848138.0, "step": 19911 }, { "epoch": 2.5330110672942374, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.11245346069336, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8696706295013428, "num_tokens": 759881877.0, "step": 19912 }, { "epoch": 2.533138277572828, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.111471176147461, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8832932114601135, "num_tokens": 759922625.0, "step": 19913 }, { "epoch": 2.5332654878514185, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.05092716217041, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8846474289894104, "num_tokens": 759952429.0, "step": 19914 }, { "epoch": 2.533392698130009, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.132654190063477, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8757548332214355, "num_tokens": 759987678.0, "step": 19915 }, { "epoch": 2.5335199084085995, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.073712348937988, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8606635332107544, "num_tokens": 760023943.0, "step": 19916 }, { "epoch": 2.53364711868719, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.183932304382324, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8762456178665161, "num_tokens": 760068229.0, "step": 19917 }, { "epoch": 2.5337743289657806, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.096433639526367, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8601976633071899, "num_tokens": 760109440.0, "step": 19918 }, { "epoch": 2.533901539244371, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.015049934387207, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8744252920150757, "num_tokens": 760146616.0, "step": 19919 }, { "epoch": 2.534028749522961, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.106062889099121, "learning_rate": 1e-06, "loss": 0.4542, "mean_token_accuracy": 0.8557097911834717, "num_tokens": 760188708.0, "step": 19920 }, { "epoch": 2.534155959801552, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.050159454345703, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8895836472511292, "num_tokens": 760231355.0, "step": 19921 }, { "epoch": 2.5342831700801423, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.119003295898438, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8608971834182739, "num_tokens": 760268316.0, "step": 19922 }, { "epoch": 2.5344103803587332, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.10392951965332, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8646243214607239, "num_tokens": 760305420.0, "step": 19923 }, { "epoch": 2.5345375906373233, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.100811958312988, "learning_rate": 1e-06, "loss": 0.3431, "mean_token_accuracy": 0.8918392658233643, "num_tokens": 760342195.0, "step": 19924 }, { "epoch": 2.534664800915914, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.093062400817871, "learning_rate": 1e-06, "loss": 0.3607, "mean_token_accuracy": 0.8869598507881165, "num_tokens": 760375394.0, "step": 19925 }, { "epoch": 2.5347920111945044, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.113327026367188, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8619977831840515, "num_tokens": 760411629.0, "step": 19926 }, { "epoch": 2.534919221473095, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.060554504394531, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8747830390930176, "num_tokens": 760456212.0, "step": 19927 }, { "epoch": 2.5350464317516854, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.113701820373535, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8641002178192139, "num_tokens": 760496385.0, "step": 19928 }, { "epoch": 2.535173642030276, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.14233112335205, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8716872930526733, "num_tokens": 760530426.0, "step": 19929 }, { "epoch": 2.5353008523088665, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.069424629211426, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8842616677284241, "num_tokens": 760568174.0, "step": 19930 }, { "epoch": 2.535428062587457, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.085171699523926, "learning_rate": 1e-06, "loss": 0.361, "mean_token_accuracy": 0.8916673064231873, "num_tokens": 760608097.0, "step": 19931 }, { "epoch": 2.5355552728660475, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.13270092010498, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8733148574829102, "num_tokens": 760643018.0, "step": 19932 }, { "epoch": 2.535682483144638, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.085739135742188, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8812161684036255, "num_tokens": 760676553.0, "step": 19933 }, { "epoch": 2.5358096934232286, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.127779960632324, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8869117498397827, "num_tokens": 760708850.0, "step": 19934 }, { "epoch": 2.535936903701819, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.03260612487793, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8695108890533447, "num_tokens": 760746088.0, "step": 19935 }, { "epoch": 2.5360641139804097, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.09357738494873, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.878728985786438, "num_tokens": 760785716.0, "step": 19936 }, { "epoch": 2.536191324259, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.04325008392334, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8824312686920166, "num_tokens": 760821301.0, "step": 19937 }, { "epoch": 2.5363185345375907, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.069561958312988, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8629299998283386, "num_tokens": 760861200.0, "step": 19938 }, { "epoch": 2.5364457448161812, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.13257884979248, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8724288940429688, "num_tokens": 760896398.0, "step": 19939 }, { "epoch": 2.5365729550947718, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.095685005187988, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8541872501373291, "num_tokens": 760937300.0, "step": 19940 }, { "epoch": 2.5367001653733623, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.09827709197998, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8681027293205261, "num_tokens": 760973726.0, "step": 19941 }, { "epoch": 2.536827375651953, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.035211563110352, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8718545436859131, "num_tokens": 761009048.0, "step": 19942 }, { "epoch": 2.5369545859305433, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.058852195739746, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8720889091491699, "num_tokens": 761050836.0, "step": 19943 }, { "epoch": 2.537081796209134, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.064667701721191, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8660551309585571, "num_tokens": 761087517.0, "step": 19944 }, { "epoch": 2.537209006487724, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.06201457977295, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8870692253112793, "num_tokens": 761121401.0, "step": 19945 }, { "epoch": 2.537336216766315, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.102621078491211, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8624252080917358, "num_tokens": 761154973.0, "step": 19946 }, { "epoch": 2.537463427044905, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.08016586303711, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8814688920974731, "num_tokens": 761194211.0, "step": 19947 }, { "epoch": 2.537590637323496, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.021844863891602, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.866426944732666, "num_tokens": 761233486.0, "step": 19948 }, { "epoch": 2.537717847602086, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.150856018066406, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8740144968032837, "num_tokens": 761275846.0, "step": 19949 }, { "epoch": 2.5378450578806766, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.055882453918457, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.863266110420227, "num_tokens": 761311210.0, "step": 19950 }, { "epoch": 2.537972268159267, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.056560516357422, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8732794523239136, "num_tokens": 761349961.0, "step": 19951 }, { "epoch": 2.5380994784378577, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.115092277526855, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8795117735862732, "num_tokens": 761386993.0, "step": 19952 }, { "epoch": 2.538226688716448, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.125722885131836, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8627746105194092, "num_tokens": 761426910.0, "step": 19953 }, { "epoch": 2.5383538989950387, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.119019508361816, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8754755258560181, "num_tokens": 761466633.0, "step": 19954 }, { "epoch": 2.5384811092736292, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.079804420471191, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8607232570648193, "num_tokens": 761497670.0, "step": 19955 }, { "epoch": 2.5386083195522198, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.156824111938477, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8694518804550171, "num_tokens": 761534626.0, "step": 19956 }, { "epoch": 2.5387355298308103, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.079004287719727, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8798660635948181, "num_tokens": 761581125.0, "step": 19957 }, { "epoch": 2.538862740109401, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.206588745117188, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8733202815055847, "num_tokens": 761616533.0, "step": 19958 }, { "epoch": 2.5389899503879914, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.094808578491211, "learning_rate": 1e-06, "loss": 0.4412, "mean_token_accuracy": 0.8635947108268738, "num_tokens": 761650615.0, "step": 19959 }, { "epoch": 2.539117160666582, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.11623477935791, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8841432332992554, "num_tokens": 761687020.0, "step": 19960 }, { "epoch": 2.5392443709451724, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.159088134765625, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8884756565093994, "num_tokens": 761721653.0, "step": 19961 }, { "epoch": 2.539371581223763, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.119173049926758, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8815433979034424, "num_tokens": 761758377.0, "step": 19962 }, { "epoch": 2.5394987915023535, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.18077564239502, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.872085452079773, "num_tokens": 761795263.0, "step": 19963 }, { "epoch": 2.539626001780944, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.06912612915039, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8832213878631592, "num_tokens": 761834758.0, "step": 19964 }, { "epoch": 2.5397532120595345, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.084065437316895, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8663035035133362, "num_tokens": 761880482.0, "step": 19965 }, { "epoch": 2.539880422338125, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.079010963439941, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8719297647476196, "num_tokens": 761922494.0, "step": 19966 }, { "epoch": 2.5400076326167156, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.191669464111328, "learning_rate": 1e-06, "loss": 0.3461, "mean_token_accuracy": 0.8908444046974182, "num_tokens": 761959462.0, "step": 19967 }, { "epoch": 2.5401348428953057, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.127771377563477, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8639326095581055, "num_tokens": 761990316.0, "step": 19968 }, { "epoch": 2.5402620531738966, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.093931198120117, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.867597758769989, "num_tokens": 762023671.0, "step": 19969 }, { "epoch": 2.5403892634524867, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.056929588317871, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8785926699638367, "num_tokens": 762056719.0, "step": 19970 }, { "epoch": 2.5405164737310777, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.073999404907227, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8743305206298828, "num_tokens": 762098802.0, "step": 19971 }, { "epoch": 2.540643684009668, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.137146949768066, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8759434819221497, "num_tokens": 762134470.0, "step": 19972 }, { "epoch": 2.5407708942882588, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.036662101745605, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.875367283821106, "num_tokens": 762177163.0, "step": 19973 }, { "epoch": 2.540898104566849, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.079244613647461, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8813213109970093, "num_tokens": 762217323.0, "step": 19974 }, { "epoch": 2.5410253148454394, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.156235694885254, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8695366978645325, "num_tokens": 762252144.0, "step": 19975 }, { "epoch": 2.54115252512403, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.127448081970215, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8664323091506958, "num_tokens": 762282108.0, "step": 19976 }, { "epoch": 2.5412797354026204, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.167925834655762, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8762434720993042, "num_tokens": 762318154.0, "step": 19977 }, { "epoch": 2.541406945681211, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.107881546020508, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8705488443374634, "num_tokens": 762354704.0, "step": 19978 }, { "epoch": 2.5415341559598015, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.125717163085938, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8752080798149109, "num_tokens": 762389943.0, "step": 19979 }, { "epoch": 2.541661366238392, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.132332801818848, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.8761622905731201, "num_tokens": 762421560.0, "step": 19980 }, { "epoch": 2.5417885765169825, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.103418350219727, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8788152933120728, "num_tokens": 762458509.0, "step": 19981 }, { "epoch": 2.541915786795573, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.074014663696289, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.873378574848175, "num_tokens": 762502382.0, "step": 19982 }, { "epoch": 2.5420429970741636, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.092680931091309, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8690861463546753, "num_tokens": 762536611.0, "step": 19983 }, { "epoch": 2.542170207352754, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.130167961120605, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8737033605575562, "num_tokens": 762577801.0, "step": 19984 }, { "epoch": 2.5422974176313446, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.102755546569824, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8752228021621704, "num_tokens": 762617413.0, "step": 19985 }, { "epoch": 2.542424627909935, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.119823455810547, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8821761012077332, "num_tokens": 762654811.0, "step": 19986 }, { "epoch": 2.5425518381885257, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.10952091217041, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8789127469062805, "num_tokens": 762692902.0, "step": 19987 }, { "epoch": 2.5426790484671162, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.16746997833252, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8742873668670654, "num_tokens": 762731984.0, "step": 19988 }, { "epoch": 2.5428062587457068, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.10128402709961, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8821732997894287, "num_tokens": 762767104.0, "step": 19989 }, { "epoch": 2.5429334690242973, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.101612091064453, "learning_rate": 1e-06, "loss": 0.4292, "mean_token_accuracy": 0.8618434071540833, "num_tokens": 762804982.0, "step": 19990 }, { "epoch": 2.543060679302888, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.112215042114258, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8633979558944702, "num_tokens": 762842956.0, "step": 19991 }, { "epoch": 2.5431878895814783, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.15050220489502, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8795685768127441, "num_tokens": 762875445.0, "step": 19992 }, { "epoch": 2.5433150998600684, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18262004852295, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8870208263397217, "num_tokens": 762906061.0, "step": 19993 }, { "epoch": 2.5434423101386594, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.054727554321289, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8818488121032715, "num_tokens": 762945550.0, "step": 19994 }, { "epoch": 2.5435695204172495, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.077678680419922, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.873175859451294, "num_tokens": 762982586.0, "step": 19995 }, { "epoch": 2.5436967306958405, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.161252975463867, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8748347163200378, "num_tokens": 763013527.0, "step": 19996 }, { "epoch": 2.5438239409744305, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.11013412475586, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8683215379714966, "num_tokens": 763054363.0, "step": 19997 }, { "epoch": 2.5439511512530215, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.15703296661377, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8689550161361694, "num_tokens": 763096834.0, "step": 19998 }, { "epoch": 2.5440783615316116, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.131046295166016, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8727195858955383, "num_tokens": 763139611.0, "step": 19999 }, { "epoch": 2.544205571810202, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.111128807067871, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8741357326507568, "num_tokens": 763181085.0, "step": 20000 }, { "epoch": 2.5443327820887927, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.144155502319336, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8710673451423645, "num_tokens": 763219811.0, "step": 20001 }, { "epoch": 2.544459992367383, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.108490943908691, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8624695539474487, "num_tokens": 763257745.0, "step": 20002 }, { "epoch": 2.5445872026459737, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.130733489990234, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8573890328407288, "num_tokens": 763299475.0, "step": 20003 }, { "epoch": 2.5447144129245642, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.120222091674805, "learning_rate": 1e-06, "loss": 0.4634, "mean_token_accuracy": 0.8519376516342163, "num_tokens": 763333960.0, "step": 20004 }, { "epoch": 2.5448416232031548, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.13587760925293, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8641030788421631, "num_tokens": 763378646.0, "step": 20005 }, { "epoch": 2.5449688334817453, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.137011528015137, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8773752450942993, "num_tokens": 763412441.0, "step": 20006 }, { "epoch": 2.545096043760336, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.184803009033203, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.877385139465332, "num_tokens": 763446713.0, "step": 20007 }, { "epoch": 2.5452232540389264, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.15092658996582, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8827933073043823, "num_tokens": 763478686.0, "step": 20008 }, { "epoch": 2.545350464317517, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.120172500610352, "learning_rate": 1e-06, "loss": 0.3573, "mean_token_accuracy": 0.8879460096359253, "num_tokens": 763517258.0, "step": 20009 }, { "epoch": 2.5454776745961074, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.12324047088623, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8682976961135864, "num_tokens": 763553229.0, "step": 20010 }, { "epoch": 2.545604884874698, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.175153732299805, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8806427121162415, "num_tokens": 763594274.0, "step": 20011 }, { "epoch": 2.5457320951532885, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.105733871459961, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.865312933921814, "num_tokens": 763632162.0, "step": 20012 }, { "epoch": 2.545859305431879, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.150739669799805, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.874245822429657, "num_tokens": 763666797.0, "step": 20013 }, { "epoch": 2.5459865157104695, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.14079475402832, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8719245195388794, "num_tokens": 763705475.0, "step": 20014 }, { "epoch": 2.54611372598906, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.116875648498535, "learning_rate": 1e-06, "loss": 0.3503, "mean_token_accuracy": 0.8894877433776855, "num_tokens": 763739434.0, "step": 20015 }, { "epoch": 2.5462409362676506, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.166349411010742, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8617469072341919, "num_tokens": 763773262.0, "step": 20016 }, { "epoch": 2.546368146546241, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.142451286315918, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8616209030151367, "num_tokens": 763812573.0, "step": 20017 }, { "epoch": 2.546495356824831, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.060113906860352, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8760284185409546, "num_tokens": 763848555.0, "step": 20018 }, { "epoch": 2.546622567103422, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.213616371154785, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8717942237854004, "num_tokens": 763884608.0, "step": 20019 }, { "epoch": 2.5467497773820122, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.105874061584473, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8690569400787354, "num_tokens": 763922038.0, "step": 20020 }, { "epoch": 2.546876987660603, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.077127456665039, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8697243928909302, "num_tokens": 763958089.0, "step": 20021 }, { "epoch": 2.5470041979391933, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.20916748046875, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8741158843040466, "num_tokens": 764000802.0, "step": 20022 }, { "epoch": 2.547131408217784, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.098687171936035, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8779605031013489, "num_tokens": 764036628.0, "step": 20023 }, { "epoch": 2.5472586184963744, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.197385787963867, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8754357099533081, "num_tokens": 764073643.0, "step": 20024 }, { "epoch": 2.547385828774965, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.148834228515625, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8793195486068726, "num_tokens": 764110473.0, "step": 20025 }, { "epoch": 2.5475130390535554, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.143381118774414, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8724957704544067, "num_tokens": 764142941.0, "step": 20026 }, { "epoch": 2.547640249332146, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.17076301574707, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8763649463653564, "num_tokens": 764178752.0, "step": 20027 }, { "epoch": 2.5477674596107365, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.212505340576172, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8707185983657837, "num_tokens": 764217440.0, "step": 20028 }, { "epoch": 2.547894669889327, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.144586563110352, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8649554252624512, "num_tokens": 764257524.0, "step": 20029 }, { "epoch": 2.5480218801679175, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.167678833007812, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8686132431030273, "num_tokens": 764293645.0, "step": 20030 }, { "epoch": 2.548149090446508, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.183753967285156, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8804752826690674, "num_tokens": 764331660.0, "step": 20031 }, { "epoch": 2.5482763007250986, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.131739616394043, "learning_rate": 1e-06, "loss": 0.3455, "mean_token_accuracy": 0.8906692862510681, "num_tokens": 764376577.0, "step": 20032 }, { "epoch": 2.548403511003689, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.175127983093262, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8711135983467102, "num_tokens": 764415492.0, "step": 20033 }, { "epoch": 2.5485307212822796, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.142913818359375, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8635462522506714, "num_tokens": 764453948.0, "step": 20034 }, { "epoch": 2.54865793156087, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.15384292602539, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8774905204772949, "num_tokens": 764493906.0, "step": 20035 }, { "epoch": 2.5487851418394607, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.178377151489258, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8669863939285278, "num_tokens": 764532379.0, "step": 20036 }, { "epoch": 2.5489123521180512, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.120506286621094, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8560181856155396, "num_tokens": 764567726.0, "step": 20037 }, { "epoch": 2.5490395623966418, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.092460632324219, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8739815354347229, "num_tokens": 764606229.0, "step": 20038 }, { "epoch": 2.5491667726752323, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.093705177307129, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.883465051651001, "num_tokens": 764641346.0, "step": 20039 }, { "epoch": 2.549293982953823, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.116634368896484, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8747539520263672, "num_tokens": 764687286.0, "step": 20040 }, { "epoch": 2.5494211932324133, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.101953506469727, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8765774965286255, "num_tokens": 764729235.0, "step": 20041 }, { "epoch": 2.549548403511004, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.090093612670898, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8658591508865356, "num_tokens": 764771834.0, "step": 20042 }, { "epoch": 2.549675613789594, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.163228034973145, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8734400272369385, "num_tokens": 764808307.0, "step": 20043 }, { "epoch": 2.549802824068185, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.14755630493164, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8658519983291626, "num_tokens": 764848373.0, "step": 20044 }, { "epoch": 2.549930034346775, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.141596794128418, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8697559237480164, "num_tokens": 764886468.0, "step": 20045 }, { "epoch": 2.550057244625366, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.107400894165039, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8745367527008057, "num_tokens": 764928690.0, "step": 20046 }, { "epoch": 2.550184454903956, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.21560001373291, "learning_rate": 1e-06, "loss": 0.5087, "mean_token_accuracy": 0.8400643467903137, "num_tokens": 764969148.0, "step": 20047 }, { "epoch": 2.5503116651825466, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.138224601745605, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8668391704559326, "num_tokens": 765010754.0, "step": 20048 }, { "epoch": 2.550438875461137, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.13594913482666, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8767956495285034, "num_tokens": 765043562.0, "step": 20049 }, { "epoch": 2.5505660857397277, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18695068359375, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.8545587062835693, "num_tokens": 765085638.0, "step": 20050 }, { "epoch": 2.550693296018318, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.123985290527344, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8798409104347229, "num_tokens": 765115847.0, "step": 20051 }, { "epoch": 2.5508205062969087, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.185604095458984, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8718327879905701, "num_tokens": 765146839.0, "step": 20052 }, { "epoch": 2.5509477165754992, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.134775161743164, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8737311959266663, "num_tokens": 765185921.0, "step": 20053 }, { "epoch": 2.5510749268540898, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.119050979614258, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8735028505325317, "num_tokens": 765224534.0, "step": 20054 }, { "epoch": 2.5512021371326803, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.182419776916504, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8810955286026001, "num_tokens": 765267160.0, "step": 20055 }, { "epoch": 2.551329347411271, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.167936325073242, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8744540214538574, "num_tokens": 765304770.0, "step": 20056 }, { "epoch": 2.5514565576898613, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.167423248291016, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8709034323692322, "num_tokens": 765341966.0, "step": 20057 }, { "epoch": 2.551583767968452, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.066218376159668, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8883035182952881, "num_tokens": 765382653.0, "step": 20058 }, { "epoch": 2.5517109782470424, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.186505317687988, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8713294267654419, "num_tokens": 765414940.0, "step": 20059 }, { "epoch": 2.551838188525633, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.130139350891113, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8749053478240967, "num_tokens": 765445549.0, "step": 20060 }, { "epoch": 2.5519653988042235, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.142379760742188, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8807111382484436, "num_tokens": 765484474.0, "step": 20061 }, { "epoch": 2.552092609082814, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.11063289642334, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8732901811599731, "num_tokens": 765522651.0, "step": 20062 }, { "epoch": 2.5522198193614045, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.15054988861084, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8685267567634583, "num_tokens": 765565983.0, "step": 20063 }, { "epoch": 2.552347029639995, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.125304222106934, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.873596727848053, "num_tokens": 765604056.0, "step": 20064 }, { "epoch": 2.5524742399185856, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.139848709106445, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8823192715644836, "num_tokens": 765646062.0, "step": 20065 }, { "epoch": 2.5526014501971757, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.175420761108398, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8727439641952515, "num_tokens": 765687284.0, "step": 20066 }, { "epoch": 2.5527286604757666, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.192948341369629, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.860617995262146, "num_tokens": 765723743.0, "step": 20067 }, { "epoch": 2.5528558707543567, "ewc_loss": 0.037841796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.055830955505371, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8701850175857544, "num_tokens": 765768472.0, "step": 20068 }, { "epoch": 2.5529830810329477, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.164995193481445, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8885153532028198, "num_tokens": 765809040.0, "step": 20069 }, { "epoch": 2.5531102913115378, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.090459823608398, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8745678663253784, "num_tokens": 765845968.0, "step": 20070 }, { "epoch": 2.5532375015901287, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.172002792358398, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8785626888275146, "num_tokens": 765881037.0, "step": 20071 }, { "epoch": 2.553364711868719, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.08588695526123, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8784999847412109, "num_tokens": 765918945.0, "step": 20072 }, { "epoch": 2.5534919221473094, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.179333686828613, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8857196569442749, "num_tokens": 765956128.0, "step": 20073 }, { "epoch": 2.5536191324259, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.088616371154785, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.8864018321037292, "num_tokens": 765997109.0, "step": 20074 }, { "epoch": 2.5537463427044904, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.164928436279297, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8710430860519409, "num_tokens": 766032557.0, "step": 20075 }, { "epoch": 2.553873552983081, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.090620994567871, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8693376183509827, "num_tokens": 766075285.0, "step": 20076 }, { "epoch": 2.5540007632616715, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.142688751220703, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8710176348686218, "num_tokens": 766110600.0, "step": 20077 }, { "epoch": 2.554127973540262, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.108451843261719, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.862818717956543, "num_tokens": 766146877.0, "step": 20078 }, { "epoch": 2.5542551838188525, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.142802238464355, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8688961267471313, "num_tokens": 766185164.0, "step": 20079 }, { "epoch": 2.554382394097443, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.114252090454102, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8923717737197876, "num_tokens": 766221442.0, "step": 20080 }, { "epoch": 2.5545096043760336, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.14573860168457, "learning_rate": 1e-06, "loss": 0.4785, "mean_token_accuracy": 0.8510352373123169, "num_tokens": 766255258.0, "step": 20081 }, { "epoch": 2.554636814654624, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.123892784118652, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8712856769561768, "num_tokens": 766292392.0, "step": 20082 }, { "epoch": 2.5547640249332146, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.164467811584473, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8641239404678345, "num_tokens": 766335312.0, "step": 20083 }, { "epoch": 2.554891235211805, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.078553199768066, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8743060827255249, "num_tokens": 766372189.0, "step": 20084 }, { "epoch": 2.5550184454903957, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.199560165405273, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.863252580165863, "num_tokens": 766414021.0, "step": 20085 }, { "epoch": 2.5551456557689862, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.137784004211426, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8863069415092468, "num_tokens": 766452005.0, "step": 20086 }, { "epoch": 2.5552728660475768, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.140016555786133, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8672117590904236, "num_tokens": 766492150.0, "step": 20087 }, { "epoch": 2.5554000763261673, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.139936447143555, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8840292692184448, "num_tokens": 766533889.0, "step": 20088 }, { "epoch": 2.555527286604758, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.224180221557617, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8738134503364563, "num_tokens": 766565008.0, "step": 20089 }, { "epoch": 2.5556544968833483, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.131412506103516, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8740366101264954, "num_tokens": 766605744.0, "step": 20090 }, { "epoch": 2.5557817071619384, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.10452938079834, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8821158409118652, "num_tokens": 766645782.0, "step": 20091 }, { "epoch": 2.5559089174405294, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.134481430053711, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8634511828422546, "num_tokens": 766683135.0, "step": 20092 }, { "epoch": 2.5560361277191195, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.135245323181152, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8745368719100952, "num_tokens": 766720691.0, "step": 20093 }, { "epoch": 2.5561633379977104, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.14011001586914, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.868776261806488, "num_tokens": 766755426.0, "step": 20094 }, { "epoch": 2.5562905482763005, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.250506401062012, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8785287737846375, "num_tokens": 766788315.0, "step": 20095 }, { "epoch": 2.5564177585548915, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.118942260742188, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8889821767807007, "num_tokens": 766825664.0, "step": 20096 }, { "epoch": 2.5565449688334816, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.153162956237793, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8855613470077515, "num_tokens": 766868145.0, "step": 20097 }, { "epoch": 2.556672179112072, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.143028259277344, "learning_rate": 1e-06, "loss": 0.4446, "mean_token_accuracy": 0.8626555800437927, "num_tokens": 766907278.0, "step": 20098 }, { "epoch": 2.5567993893906626, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.174649238586426, "learning_rate": 1e-06, "loss": 0.3612, "mean_token_accuracy": 0.8857992887496948, "num_tokens": 766941662.0, "step": 20099 }, { "epoch": 2.556926599669253, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.173927307128906, "learning_rate": 1e-06, "loss": 0.4695, "mean_token_accuracy": 0.8558014035224915, "num_tokens": 766977725.0, "step": 20100 }, { "epoch": 2.5570538099478437, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.186285972595215, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8693445920944214, "num_tokens": 767011011.0, "step": 20101 }, { "epoch": 2.5571810202264342, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.184762954711914, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8712789416313171, "num_tokens": 767047078.0, "step": 20102 }, { "epoch": 2.5573082305050248, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.171632766723633, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8694884181022644, "num_tokens": 767089208.0, "step": 20103 }, { "epoch": 2.5574354407836153, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.149773597717285, "learning_rate": 1e-06, "loss": 0.4595, "mean_token_accuracy": 0.8577332496643066, "num_tokens": 767131225.0, "step": 20104 }, { "epoch": 2.557562651062206, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.192890167236328, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8617337346076965, "num_tokens": 767165222.0, "step": 20105 }, { "epoch": 2.5576898613407963, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.11684799194336, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8749513030052185, "num_tokens": 767200870.0, "step": 20106 }, { "epoch": 2.557817071619387, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.25613021850586, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8724409937858582, "num_tokens": 767236881.0, "step": 20107 }, { "epoch": 2.5579442818979774, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.149041175842285, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.879403293132782, "num_tokens": 767269716.0, "step": 20108 }, { "epoch": 2.558071492176568, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.181140899658203, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8640223741531372, "num_tokens": 767306840.0, "step": 20109 }, { "epoch": 2.5581987024551585, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.146551132202148, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8858890533447266, "num_tokens": 767346966.0, "step": 20110 }, { "epoch": 2.558325912733749, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.172738075256348, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8694282174110413, "num_tokens": 767387084.0, "step": 20111 }, { "epoch": 2.5584531230123395, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.227828979492188, "learning_rate": 1e-06, "loss": 0.4581, "mean_token_accuracy": 0.8545739650726318, "num_tokens": 767417857.0, "step": 20112 }, { "epoch": 2.55858033329093, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.16627311706543, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8753317594528198, "num_tokens": 767455771.0, "step": 20113 }, { "epoch": 2.5587075435695206, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.119665145874023, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8600336313247681, "num_tokens": 767493187.0, "step": 20114 }, { "epoch": 2.558834753848111, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.19238567352295, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.865200400352478, "num_tokens": 767534526.0, "step": 20115 }, { "epoch": 2.558961964126701, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.16331672668457, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8709890842437744, "num_tokens": 767569297.0, "step": 20116 }, { "epoch": 2.559089174405292, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.144556999206543, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8691756725311279, "num_tokens": 767605860.0, "step": 20117 }, { "epoch": 2.5592163846838822, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.118189811706543, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8761141300201416, "num_tokens": 767645252.0, "step": 20118 }, { "epoch": 2.559343594962473, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.157739639282227, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8740683794021606, "num_tokens": 767681942.0, "step": 20119 }, { "epoch": 2.5594708052410633, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.173135757446289, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.871838390827179, "num_tokens": 767720631.0, "step": 20120 }, { "epoch": 2.559598015519654, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.172113418579102, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8775855898857117, "num_tokens": 767757018.0, "step": 20121 }, { "epoch": 2.5597252257982444, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.168542861938477, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8768531084060669, "num_tokens": 767796046.0, "step": 20122 }, { "epoch": 2.559852436076835, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.217933654785156, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8837104439735413, "num_tokens": 767835678.0, "step": 20123 }, { "epoch": 2.5599796463554254, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.11971378326416, "learning_rate": 1e-06, "loss": 0.3599, "mean_token_accuracy": 0.8865134716033936, "num_tokens": 767870515.0, "step": 20124 }, { "epoch": 2.560106856634016, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.197664260864258, "learning_rate": 1e-06, "loss": 0.4511, "mean_token_accuracy": 0.8576568961143494, "num_tokens": 767911573.0, "step": 20125 }, { "epoch": 2.5602340669126065, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.197628021240234, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8726656436920166, "num_tokens": 767950161.0, "step": 20126 }, { "epoch": 2.560361277191197, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.12289810180664, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8700848817825317, "num_tokens": 767989431.0, "step": 20127 }, { "epoch": 2.5604884874697875, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.226218223571777, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8830657601356506, "num_tokens": 768023551.0, "step": 20128 }, { "epoch": 2.560615697748378, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.15400505065918, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8811871409416199, "num_tokens": 768055317.0, "step": 20129 }, { "epoch": 2.5607429080269686, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.308128356933594, "learning_rate": 1e-06, "loss": 0.4305, "mean_token_accuracy": 0.8662038445472717, "num_tokens": 768089575.0, "step": 20130 }, { "epoch": 2.560870118305559, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.181077003479004, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8638641834259033, "num_tokens": 768131795.0, "step": 20131 }, { "epoch": 2.5609973285841496, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.200004577636719, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8777287006378174, "num_tokens": 768165714.0, "step": 20132 }, { "epoch": 2.56112453886274, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.225677490234375, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8815093040466309, "num_tokens": 768203475.0, "step": 20133 }, { "epoch": 2.5612517491413307, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.104185104370117, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8887951970100403, "num_tokens": 768244841.0, "step": 20134 }, { "epoch": 2.561378959419921, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.205262184143066, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8685921430587769, "num_tokens": 768286502.0, "step": 20135 }, { "epoch": 2.5615061696985117, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.11896800994873, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8815518617630005, "num_tokens": 768320628.0, "step": 20136 }, { "epoch": 2.5616333799771023, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.233047485351562, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8666768074035645, "num_tokens": 768353630.0, "step": 20137 }, { "epoch": 2.561760590255693, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.141560554504395, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8732983469963074, "num_tokens": 768388170.0, "step": 20138 }, { "epoch": 2.5618878005342833, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.150955200195312, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.8892291784286499, "num_tokens": 768428238.0, "step": 20139 }, { "epoch": 2.562015010812874, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.153730392456055, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8870280981063843, "num_tokens": 768460075.0, "step": 20140 }, { "epoch": 2.562142221091464, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.195358276367188, "learning_rate": 1e-06, "loss": 0.3397, "mean_token_accuracy": 0.8903673887252808, "num_tokens": 768493982.0, "step": 20141 }, { "epoch": 2.562269431370055, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.202384948730469, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8773276209831238, "num_tokens": 768529121.0, "step": 20142 }, { "epoch": 2.562396641648645, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.227692604064941, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8742374777793884, "num_tokens": 768560730.0, "step": 20143 }, { "epoch": 2.562523851927236, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.158172607421875, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.8838320970535278, "num_tokens": 768603015.0, "step": 20144 }, { "epoch": 2.562651062205826, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.193943977355957, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8733222484588623, "num_tokens": 768639847.0, "step": 20145 }, { "epoch": 2.5627782724844166, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.1884126663208, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8708056211471558, "num_tokens": 768681614.0, "step": 20146 }, { "epoch": 2.562905482763007, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.10938835144043, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8762123584747314, "num_tokens": 768722052.0, "step": 20147 }, { "epoch": 2.5630326930415976, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.154939651489258, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8666929006576538, "num_tokens": 768754191.0, "step": 20148 }, { "epoch": 2.563159903320188, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.177711486816406, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.862706184387207, "num_tokens": 768792981.0, "step": 20149 }, { "epoch": 2.5632871135987787, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.166752815246582, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8647770881652832, "num_tokens": 768827040.0, "step": 20150 }, { "epoch": 2.5634143238773692, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.13884162902832, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8714427351951599, "num_tokens": 768865669.0, "step": 20151 }, { "epoch": 2.5635415341559598, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.203110694885254, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.872930645942688, "num_tokens": 768903306.0, "step": 20152 }, { "epoch": 2.5636687444345503, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.142477989196777, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8773474097251892, "num_tokens": 768946322.0, "step": 20153 }, { "epoch": 2.563795954713141, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.16076374053955, "learning_rate": 1e-06, "loss": 0.3399, "mean_token_accuracy": 0.8914756178855896, "num_tokens": 768987350.0, "step": 20154 }, { "epoch": 2.5639231649917313, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.112701416015625, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8821948766708374, "num_tokens": 769030762.0, "step": 20155 }, { "epoch": 2.564050375270322, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 14.13766860961914, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8848791122436523, "num_tokens": 769069827.0, "step": 20156 }, { "epoch": 2.5641775855489124, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.157562255859375, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8736927509307861, "num_tokens": 769107742.0, "step": 20157 }, { "epoch": 2.564304795827503, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.179825782775879, "learning_rate": 1e-06, "loss": 0.422, "mean_token_accuracy": 0.8719958066940308, "num_tokens": 769153286.0, "step": 20158 }, { "epoch": 2.5644320061060935, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.13652229309082, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8748780488967896, "num_tokens": 769192646.0, "step": 20159 }, { "epoch": 2.564559216384684, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.23587703704834, "learning_rate": 1e-06, "loss": 0.4675, "mean_token_accuracy": 0.8519855737686157, "num_tokens": 769230360.0, "step": 20160 }, { "epoch": 2.5646864266632745, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.129951477050781, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8889985680580139, "num_tokens": 769270754.0, "step": 20161 }, { "epoch": 2.564813636941865, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.15427303314209, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8506134748458862, "num_tokens": 769310109.0, "step": 20162 }, { "epoch": 2.5649408472204556, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.097264289855957, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8721917867660522, "num_tokens": 769352101.0, "step": 20163 }, { "epoch": 2.5650680574990457, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.170273780822754, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8875148892402649, "num_tokens": 769386570.0, "step": 20164 }, { "epoch": 2.5651952677776366, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.171623229980469, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8845654726028442, "num_tokens": 769427167.0, "step": 20165 }, { "epoch": 2.5653224780562267, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.164668083190918, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8633354902267456, "num_tokens": 769465456.0, "step": 20166 }, { "epoch": 2.5654496883348177, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.132366180419922, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.8864129781723022, "num_tokens": 769497135.0, "step": 20167 }, { "epoch": 2.5655768986134078, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.119179725646973, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8682460784912109, "num_tokens": 769533706.0, "step": 20168 }, { "epoch": 2.5657041088919987, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.162589073181152, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8685283660888672, "num_tokens": 769578698.0, "step": 20169 }, { "epoch": 2.565831319170589, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.09312915802002, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8858327865600586, "num_tokens": 769619799.0, "step": 20170 }, { "epoch": 2.5659585294491793, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.17298698425293, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8735218644142151, "num_tokens": 769655973.0, "step": 20171 }, { "epoch": 2.56608573972777, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.111699104309082, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8693611025810242, "num_tokens": 769692333.0, "step": 20172 }, { "epoch": 2.5662129500063604, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.176474571228027, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8775322437286377, "num_tokens": 769731804.0, "step": 20173 }, { "epoch": 2.566340160284951, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.165182113647461, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8792614340782166, "num_tokens": 769767409.0, "step": 20174 }, { "epoch": 2.5664673705635415, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.115129470825195, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8712627291679382, "num_tokens": 769806468.0, "step": 20175 }, { "epoch": 2.566594580842132, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.193798065185547, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.881859302520752, "num_tokens": 769842025.0, "step": 20176 }, { "epoch": 2.5667217911207225, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.213027000427246, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8687705993652344, "num_tokens": 769884125.0, "step": 20177 }, { "epoch": 2.566849001399313, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18531608581543, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8629192113876343, "num_tokens": 769920523.0, "step": 20178 }, { "epoch": 2.5669762116779036, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.181227684020996, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8715689182281494, "num_tokens": 769960141.0, "step": 20179 }, { "epoch": 2.567103421956494, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.223401069641113, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8753507137298584, "num_tokens": 769995281.0, "step": 20180 }, { "epoch": 2.5672306322350846, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.198973655700684, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8775084614753723, "num_tokens": 770028230.0, "step": 20181 }, { "epoch": 2.567357842513675, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.132341384887695, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8808802366256714, "num_tokens": 770068996.0, "step": 20182 }, { "epoch": 2.5674850527922657, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.190495491027832, "learning_rate": 1e-06, "loss": 0.4315, "mean_token_accuracy": 0.8642079830169678, "num_tokens": 770107018.0, "step": 20183 }, { "epoch": 2.567612263070856, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.182244300842285, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8724365830421448, "num_tokens": 770140251.0, "step": 20184 }, { "epoch": 2.5677394733494467, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.19630241394043, "learning_rate": 1e-06, "loss": 0.4403, "mean_token_accuracy": 0.8663053512573242, "num_tokens": 770182241.0, "step": 20185 }, { "epoch": 2.5678666836280373, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.203997611999512, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8870270252227783, "num_tokens": 770223267.0, "step": 20186 }, { "epoch": 2.567993893906628, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.165971755981445, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8631260395050049, "num_tokens": 770261856.0, "step": 20187 }, { "epoch": 2.5681211041852183, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.251343727111816, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8631475567817688, "num_tokens": 770299762.0, "step": 20188 }, { "epoch": 2.5682483144638084, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.165812492370605, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8687595129013062, "num_tokens": 770333690.0, "step": 20189 }, { "epoch": 2.5683755247423994, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 14.209266662597656, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.87943434715271, "num_tokens": 770375755.0, "step": 20190 }, { "epoch": 2.5685027350209895, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.121337890625, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8830115795135498, "num_tokens": 770417805.0, "step": 20191 }, { "epoch": 2.5686299452995804, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.269512176513672, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8723535537719727, "num_tokens": 770458183.0, "step": 20192 }, { "epoch": 2.5687571555781705, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.181446075439453, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8769216537475586, "num_tokens": 770488973.0, "step": 20193 }, { "epoch": 2.5688843658567615, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.213428497314453, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8681620955467224, "num_tokens": 770529030.0, "step": 20194 }, { "epoch": 2.5690115761353516, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.160759925842285, "learning_rate": 1e-06, "loss": 0.436, "mean_token_accuracy": 0.8631816506385803, "num_tokens": 770574697.0, "step": 20195 }, { "epoch": 2.569138786413942, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.192915916442871, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8796105980873108, "num_tokens": 770615417.0, "step": 20196 }, { "epoch": 2.5692659966925326, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.16491413116455, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.860499918460846, "num_tokens": 770653228.0, "step": 20197 }, { "epoch": 2.569393206971123, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.161388397216797, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8876529335975647, "num_tokens": 770693876.0, "step": 20198 }, { "epoch": 2.5695204172497137, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.152552604675293, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8723791837692261, "num_tokens": 770728754.0, "step": 20199 }, { "epoch": 2.5696476275283042, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.153199195861816, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8714166879653931, "num_tokens": 770767961.0, "step": 20200 }, { "epoch": 2.5697748378068948, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.200977325439453, "learning_rate": 1e-06, "loss": 0.4917, "mean_token_accuracy": 0.8470689654350281, "num_tokens": 770808880.0, "step": 20201 }, { "epoch": 2.5699020480854853, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.238862037658691, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8677767515182495, "num_tokens": 770846483.0, "step": 20202 }, { "epoch": 2.570029258364076, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.145151138305664, "learning_rate": 1e-06, "loss": 0.476, "mean_token_accuracy": 0.854057252407074, "num_tokens": 770884234.0, "step": 20203 }, { "epoch": 2.5701564686426663, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.19528579711914, "learning_rate": 1e-06, "loss": 0.4473, "mean_token_accuracy": 0.8595603108406067, "num_tokens": 770918785.0, "step": 20204 }, { "epoch": 2.570283678921257, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.166801452636719, "learning_rate": 1e-06, "loss": 0.3355, "mean_token_accuracy": 0.8943405747413635, "num_tokens": 770956889.0, "step": 20205 }, { "epoch": 2.5704108891998474, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.205765724182129, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8600099086761475, "num_tokens": 770994937.0, "step": 20206 }, { "epoch": 2.570538099478438, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.148271560668945, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.882319986820221, "num_tokens": 771031451.0, "step": 20207 }, { "epoch": 2.5706653097570284, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18255615234375, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8807327747344971, "num_tokens": 771061575.0, "step": 20208 }, { "epoch": 2.570792520035619, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.125614166259766, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8721661567687988, "num_tokens": 771103366.0, "step": 20209 }, { "epoch": 2.5709197303142095, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.115270614624023, "learning_rate": 1e-06, "loss": 0.3894, "mean_token_accuracy": 0.8773285150527954, "num_tokens": 771142178.0, "step": 20210 }, { "epoch": 2.5710469405928, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.1896390914917, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8741168975830078, "num_tokens": 771179296.0, "step": 20211 }, { "epoch": 2.5711741508713906, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.070647239685059, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8664877414703369, "num_tokens": 771219084.0, "step": 20212 }, { "epoch": 2.571301361149981, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.141091346740723, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8798462152481079, "num_tokens": 771250609.0, "step": 20213 }, { "epoch": 2.571428571428571, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.187983512878418, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8722705841064453, "num_tokens": 771286191.0, "step": 20214 }, { "epoch": 2.571555781707162, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.156630516052246, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8715572953224182, "num_tokens": 771320898.0, "step": 20215 }, { "epoch": 2.5716829919857522, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.146751403808594, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8640700578689575, "num_tokens": 771353977.0, "step": 20216 }, { "epoch": 2.571810202264343, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.182817459106445, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8787490129470825, "num_tokens": 771391960.0, "step": 20217 }, { "epoch": 2.5719374125429333, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.183313369750977, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8659670352935791, "num_tokens": 771427764.0, "step": 20218 }, { "epoch": 2.572064622821524, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.2432279586792, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8732343912124634, "num_tokens": 771469706.0, "step": 20219 }, { "epoch": 2.5721918331001143, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.2042875289917, "learning_rate": 1e-06, "loss": 0.3359, "mean_token_accuracy": 0.8938254117965698, "num_tokens": 771504296.0, "step": 20220 }, { "epoch": 2.572319043378705, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.161248207092285, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8737062215805054, "num_tokens": 771544496.0, "step": 20221 }, { "epoch": 2.5724462536572954, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.192914962768555, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8815025687217712, "num_tokens": 771584322.0, "step": 20222 }, { "epoch": 2.572573463935886, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.198999404907227, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8822759389877319, "num_tokens": 771621424.0, "step": 20223 }, { "epoch": 2.5727006742144765, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.205446243286133, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8752435445785522, "num_tokens": 771658173.0, "step": 20224 }, { "epoch": 2.572827884493067, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.141338348388672, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8659279346466064, "num_tokens": 771698182.0, "step": 20225 }, { "epoch": 2.5729550947716575, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.237257957458496, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8577879667282104, "num_tokens": 771731559.0, "step": 20226 }, { "epoch": 2.573082305050248, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.150672912597656, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8691269159317017, "num_tokens": 771774362.0, "step": 20227 }, { "epoch": 2.5732095153288386, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.23204517364502, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8627512454986572, "num_tokens": 771810540.0, "step": 20228 }, { "epoch": 2.573336725607429, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.17245864868164, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.878063440322876, "num_tokens": 771846221.0, "step": 20229 }, { "epoch": 2.5734639358860196, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.168432235717773, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8834189772605896, "num_tokens": 771882268.0, "step": 20230 }, { "epoch": 2.57359114616461, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.25491714477539, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8729860186576843, "num_tokens": 771913176.0, "step": 20231 }, { "epoch": 2.5737183564432007, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.29440975189209, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8681716918945312, "num_tokens": 771948163.0, "step": 20232 }, { "epoch": 2.573845566721791, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.146857261657715, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.885625958442688, "num_tokens": 771988296.0, "step": 20233 }, { "epoch": 2.5739727770003817, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.179139137268066, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8807929754257202, "num_tokens": 772024330.0, "step": 20234 }, { "epoch": 2.5740999872789723, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.181751251220703, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.861330509185791, "num_tokens": 772062291.0, "step": 20235 }, { "epoch": 2.574227197557563, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.147687911987305, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8744632601737976, "num_tokens": 772110239.0, "step": 20236 }, { "epoch": 2.5743544078361533, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.109859466552734, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8688320517539978, "num_tokens": 772153201.0, "step": 20237 }, { "epoch": 2.574481618114744, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18673324584961, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8752279281616211, "num_tokens": 772186816.0, "step": 20238 }, { "epoch": 2.574608828393334, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.13081169128418, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8835004568099976, "num_tokens": 772224238.0, "step": 20239 }, { "epoch": 2.574736038671925, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.13929557800293, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8817359209060669, "num_tokens": 772263505.0, "step": 20240 }, { "epoch": 2.574863248950515, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.155110359191895, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8819743990898132, "num_tokens": 772303205.0, "step": 20241 }, { "epoch": 2.574990459229106, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.206094741821289, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8786433935165405, "num_tokens": 772344495.0, "step": 20242 }, { "epoch": 2.575117669507696, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.11972427368164, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8681687116622925, "num_tokens": 772380296.0, "step": 20243 }, { "epoch": 2.5752448797862866, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.187248229980469, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8864173889160156, "num_tokens": 772423933.0, "step": 20244 }, { "epoch": 2.575372090064877, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.167839050292969, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8601049184799194, "num_tokens": 772464514.0, "step": 20245 }, { "epoch": 2.5754993003434676, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.147420883178711, "learning_rate": 1e-06, "loss": 0.3352, "mean_token_accuracy": 0.8917544484138489, "num_tokens": 772498501.0, "step": 20246 }, { "epoch": 2.575626510622058, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.19493579864502, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8703272342681885, "num_tokens": 772538072.0, "step": 20247 }, { "epoch": 2.5757537209006487, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.179909706115723, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8738836050033569, "num_tokens": 772577434.0, "step": 20248 }, { "epoch": 2.575880931179239, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.130096435546875, "learning_rate": 1e-06, "loss": 0.3498, "mean_token_accuracy": 0.8873884081840515, "num_tokens": 772612917.0, "step": 20249 }, { "epoch": 2.5760081414578297, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18774127960205, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8851684331893921, "num_tokens": 772652939.0, "step": 20250 }, { "epoch": 2.5761353517364203, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.167121887207031, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8800598382949829, "num_tokens": 772688018.0, "step": 20251 }, { "epoch": 2.576262562015011, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.165799140930176, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8640198707580566, "num_tokens": 772728071.0, "step": 20252 }, { "epoch": 2.5763897722936013, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.227782249450684, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8753156065940857, "num_tokens": 772762730.0, "step": 20253 }, { "epoch": 2.576516982572192, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.15246295928955, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8771703243255615, "num_tokens": 772799125.0, "step": 20254 }, { "epoch": 2.5766441928507824, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.155866622924805, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8865114450454712, "num_tokens": 772838657.0, "step": 20255 }, { "epoch": 2.576771403129373, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.217684745788574, "learning_rate": 1e-06, "loss": 0.3793, "mean_token_accuracy": 0.8812379837036133, "num_tokens": 772876470.0, "step": 20256 }, { "epoch": 2.5768986134079634, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.147557258605957, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8675316572189331, "num_tokens": 772910369.0, "step": 20257 }, { "epoch": 2.577025823686554, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.202641487121582, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8786014318466187, "num_tokens": 772948129.0, "step": 20258 }, { "epoch": 2.5771530339651445, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.188435554504395, "learning_rate": 1e-06, "loss": 0.3212, "mean_token_accuracy": 0.9004076719284058, "num_tokens": 772985359.0, "step": 20259 }, { "epoch": 2.577280244243735, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.193367958068848, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8756765127182007, "num_tokens": 773019402.0, "step": 20260 }, { "epoch": 2.5774074545223256, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.17082691192627, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8736782670021057, "num_tokens": 773059176.0, "step": 20261 }, { "epoch": 2.5775346648009156, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.230496406555176, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8762558698654175, "num_tokens": 773094274.0, "step": 20262 }, { "epoch": 2.5776618750795066, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.179938316345215, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8732374310493469, "num_tokens": 773135679.0, "step": 20263 }, { "epoch": 2.5777890853580967, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.102877616882324, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.870543360710144, "num_tokens": 773174270.0, "step": 20264 }, { "epoch": 2.5779162956366877, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.128129005432129, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8754937648773193, "num_tokens": 773213582.0, "step": 20265 }, { "epoch": 2.5780435059152778, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.227339744567871, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8708141446113586, "num_tokens": 773244980.0, "step": 20266 }, { "epoch": 2.5781707161938687, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.075191497802734, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8875706195831299, "num_tokens": 773287395.0, "step": 20267 }, { "epoch": 2.578297926472459, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.121203422546387, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8705474138259888, "num_tokens": 773324593.0, "step": 20268 }, { "epoch": 2.5784251367510493, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.182379722595215, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8697465658187866, "num_tokens": 773361802.0, "step": 20269 }, { "epoch": 2.57855234702964, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.098602294921875, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8784732818603516, "num_tokens": 773401348.0, "step": 20270 }, { "epoch": 2.5786795573082304, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.17507553100586, "learning_rate": 1e-06, "loss": 0.355, "mean_token_accuracy": 0.8899959325790405, "num_tokens": 773439040.0, "step": 20271 }, { "epoch": 2.578806767586821, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.11954402923584, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8859421610832214, "num_tokens": 773471382.0, "step": 20272 }, { "epoch": 2.5789339778654115, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.181541442871094, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8637287020683289, "num_tokens": 773508924.0, "step": 20273 }, { "epoch": 2.579061188144002, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.11027717590332, "learning_rate": 1e-06, "loss": 0.3481, "mean_token_accuracy": 0.8884280323982239, "num_tokens": 773550219.0, "step": 20274 }, { "epoch": 2.5791883984225925, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.118807792663574, "learning_rate": 1e-06, "loss": 0.4155, "mean_token_accuracy": 0.8694053292274475, "num_tokens": 773589363.0, "step": 20275 }, { "epoch": 2.579315608701183, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.173751831054688, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8736143112182617, "num_tokens": 773630289.0, "step": 20276 }, { "epoch": 2.5794428189797736, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.127720832824707, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.874703586101532, "num_tokens": 773668660.0, "step": 20277 }, { "epoch": 2.579570029258364, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.137848854064941, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8712406754493713, "num_tokens": 773704318.0, "step": 20278 }, { "epoch": 2.5796972395369546, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.21179485321045, "learning_rate": 1e-06, "loss": 0.4488, "mean_token_accuracy": 0.8589121699333191, "num_tokens": 773745821.0, "step": 20279 }, { "epoch": 2.579824449815545, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.182747840881348, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8832359313964844, "num_tokens": 773782880.0, "step": 20280 }, { "epoch": 2.5799516600941357, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.155132293701172, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8606802821159363, "num_tokens": 773824880.0, "step": 20281 }, { "epoch": 2.580078870372726, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.202431678771973, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8778791427612305, "num_tokens": 773863210.0, "step": 20282 }, { "epoch": 2.5802060806513167, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.132221221923828, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8813024759292603, "num_tokens": 773903432.0, "step": 20283 }, { "epoch": 2.5803332909299073, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.181493759155273, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8826723694801331, "num_tokens": 773942424.0, "step": 20284 }, { "epoch": 2.580460501208498, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.164629936218262, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8736369013786316, "num_tokens": 773976598.0, "step": 20285 }, { "epoch": 2.5805877114870883, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.152631759643555, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8679633140563965, "num_tokens": 774016696.0, "step": 20286 }, { "epoch": 2.5807149217656784, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.1614408493042, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8681671619415283, "num_tokens": 774059046.0, "step": 20287 }, { "epoch": 2.5808421320442694, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.182913780212402, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8704948425292969, "num_tokens": 774095322.0, "step": 20288 }, { "epoch": 2.5809693423228595, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.120365142822266, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8631519079208374, "num_tokens": 774134506.0, "step": 20289 }, { "epoch": 2.5810965526014504, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.15212345123291, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8888864517211914, "num_tokens": 774176997.0, "step": 20290 }, { "epoch": 2.5812237628800405, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.185014724731445, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8693253993988037, "num_tokens": 774213900.0, "step": 20291 }, { "epoch": 2.5813509731586315, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.120222091674805, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8666974306106567, "num_tokens": 774259687.0, "step": 20292 }, { "epoch": 2.5814781834372216, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.226203918457031, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8783708214759827, "num_tokens": 774297200.0, "step": 20293 }, { "epoch": 2.581605393715812, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.15273666381836, "learning_rate": 1e-06, "loss": 0.3581, "mean_token_accuracy": 0.8869361877441406, "num_tokens": 774336642.0, "step": 20294 }, { "epoch": 2.5817326039944026, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.221638679504395, "learning_rate": 1e-06, "loss": 0.4402, "mean_token_accuracy": 0.8584292531013489, "num_tokens": 774376762.0, "step": 20295 }, { "epoch": 2.581859814272993, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.137084007263184, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8819587230682373, "num_tokens": 774421360.0, "step": 20296 }, { "epoch": 2.5819870245515837, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.194528579711914, "learning_rate": 1e-06, "loss": 0.3728, "mean_token_accuracy": 0.8815228343009949, "num_tokens": 774460850.0, "step": 20297 }, { "epoch": 2.582114234830174, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.142191886901855, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8715223073959351, "num_tokens": 774497542.0, "step": 20298 }, { "epoch": 2.5822414451087647, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.235366821289062, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8666352033615112, "num_tokens": 774538164.0, "step": 20299 }, { "epoch": 2.5823686553873553, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.166746139526367, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8689618110656738, "num_tokens": 774576118.0, "step": 20300 }, { "epoch": 2.582495865665946, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.167280197143555, "learning_rate": 1e-06, "loss": 0.3381, "mean_token_accuracy": 0.8949564099311829, "num_tokens": 774617506.0, "step": 20301 }, { "epoch": 2.5826230759445363, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.166576385498047, "learning_rate": 1e-06, "loss": 0.4779, "mean_token_accuracy": 0.8504700660705566, "num_tokens": 774653720.0, "step": 20302 }, { "epoch": 2.582750286223127, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.15893840789795, "learning_rate": 1e-06, "loss": 0.3317, "mean_token_accuracy": 0.8942129611968994, "num_tokens": 774691056.0, "step": 20303 }, { "epoch": 2.5828774965017174, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.13862133026123, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8679808378219604, "num_tokens": 774726514.0, "step": 20304 }, { "epoch": 2.583004706780308, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.195691108703613, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8659114241600037, "num_tokens": 774760831.0, "step": 20305 }, { "epoch": 2.5831319170588984, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.143793106079102, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.883576512336731, "num_tokens": 774799399.0, "step": 20306 }, { "epoch": 2.583259127337489, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.135665893554688, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8660314083099365, "num_tokens": 774839910.0, "step": 20307 }, { "epoch": 2.5833863376160795, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.22933292388916, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8741601705551147, "num_tokens": 774875694.0, "step": 20308 }, { "epoch": 2.58351354789467, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.174247741699219, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.876261830329895, "num_tokens": 774908020.0, "step": 20309 }, { "epoch": 2.5836407581732606, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.19990348815918, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.864159345626831, "num_tokens": 774946716.0, "step": 20310 }, { "epoch": 2.583767968451851, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.180310249328613, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8685367107391357, "num_tokens": 774981732.0, "step": 20311 }, { "epoch": 2.583895178730441, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.253012657165527, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8795763254165649, "num_tokens": 775015280.0, "step": 20312 }, { "epoch": 2.584022389009032, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.20715618133545, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8819230794906616, "num_tokens": 775054518.0, "step": 20313 }, { "epoch": 2.5841495992876222, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.165690422058105, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8805429339408875, "num_tokens": 775091548.0, "step": 20314 }, { "epoch": 2.584276809566213, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.178054809570312, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8621906638145447, "num_tokens": 775130916.0, "step": 20315 }, { "epoch": 2.5844040198448033, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.166629791259766, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8863003253936768, "num_tokens": 775165875.0, "step": 20316 }, { "epoch": 2.584531230123394, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.223580360412598, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8606100678443909, "num_tokens": 775211463.0, "step": 20317 }, { "epoch": 2.5846584404019843, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.150463104248047, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8826805949211121, "num_tokens": 775248338.0, "step": 20318 }, { "epoch": 2.584785650680575, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.209473609924316, "learning_rate": 1e-06, "loss": 0.4333, "mean_token_accuracy": 0.8633809685707092, "num_tokens": 775285707.0, "step": 20319 }, { "epoch": 2.5849128609591654, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.201767921447754, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8827099800109863, "num_tokens": 775322682.0, "step": 20320 }, { "epoch": 2.585040071237756, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.172959327697754, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8837124109268188, "num_tokens": 775358274.0, "step": 20321 }, { "epoch": 2.5851672815163464, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.22409439086914, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8694391250610352, "num_tokens": 775394472.0, "step": 20322 }, { "epoch": 2.585294491794937, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.16919994354248, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8841601014137268, "num_tokens": 775433647.0, "step": 20323 }, { "epoch": 2.5854217020735275, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.22163200378418, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8656238317489624, "num_tokens": 775469159.0, "step": 20324 }, { "epoch": 2.585548912352118, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.235223770141602, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8819296956062317, "num_tokens": 775507329.0, "step": 20325 }, { "epoch": 2.5856761226307086, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.264391899108887, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8720389604568481, "num_tokens": 775544638.0, "step": 20326 }, { "epoch": 2.585803332909299, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.210951805114746, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8762484788894653, "num_tokens": 775584918.0, "step": 20327 }, { "epoch": 2.5859305431878896, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.145527839660645, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8697313666343689, "num_tokens": 775624214.0, "step": 20328 }, { "epoch": 2.58605775346648, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.22885799407959, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8624860048294067, "num_tokens": 775668372.0, "step": 20329 }, { "epoch": 2.5861849637450707, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.2482271194458, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8469727039337158, "num_tokens": 775700578.0, "step": 20330 }, { "epoch": 2.586312174023661, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.173267364501953, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8798593282699585, "num_tokens": 775746980.0, "step": 20331 }, { "epoch": 2.5864393843022517, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.138461112976074, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8684076070785522, "num_tokens": 775780762.0, "step": 20332 }, { "epoch": 2.5865665945808423, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.169736862182617, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8807457685470581, "num_tokens": 775815591.0, "step": 20333 }, { "epoch": 2.586693804859433, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.15162467956543, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8660274744033813, "num_tokens": 775856414.0, "step": 20334 }, { "epoch": 2.5868210151380233, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.217653274536133, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8749189376831055, "num_tokens": 775896939.0, "step": 20335 }, { "epoch": 2.586948225416614, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.144627571105957, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8686857223510742, "num_tokens": 775941778.0, "step": 20336 }, { "epoch": 2.587075435695204, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.149572372436523, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8736810684204102, "num_tokens": 775979346.0, "step": 20337 }, { "epoch": 2.587202645973795, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.193098068237305, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8723044991493225, "num_tokens": 776013793.0, "step": 20338 }, { "epoch": 2.587329856252385, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.179383277893066, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8860036134719849, "num_tokens": 776058513.0, "step": 20339 }, { "epoch": 2.587457066530976, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.15626335144043, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8748682141304016, "num_tokens": 776102790.0, "step": 20340 }, { "epoch": 2.587584276809566, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.182316780090332, "learning_rate": 1e-06, "loss": 0.348, "mean_token_accuracy": 0.8896481990814209, "num_tokens": 776133898.0, "step": 20341 }, { "epoch": 2.5877114870881566, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.180429458618164, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8729402422904968, "num_tokens": 776167590.0, "step": 20342 }, { "epoch": 2.587838697366747, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.19008731842041, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8684412240982056, "num_tokens": 776207966.0, "step": 20343 }, { "epoch": 2.5879659076453376, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.16380500793457, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8740993738174438, "num_tokens": 776245601.0, "step": 20344 }, { "epoch": 2.588093117923928, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.168944358825684, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.8818762302398682, "num_tokens": 776285377.0, "step": 20345 }, { "epoch": 2.5882203282025187, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.218405723571777, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8646395206451416, "num_tokens": 776328116.0, "step": 20346 }, { "epoch": 2.588347538481109, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.158646583557129, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8584613800048828, "num_tokens": 776369008.0, "step": 20347 }, { "epoch": 2.5884747487596997, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.189629554748535, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8652622103691101, "num_tokens": 776410151.0, "step": 20348 }, { "epoch": 2.5886019590382903, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.147963523864746, "learning_rate": 1e-06, "loss": 0.3404, "mean_token_accuracy": 0.8975526094436646, "num_tokens": 776445944.0, "step": 20349 }, { "epoch": 2.588729169316881, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.208348274230957, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8779894113540649, "num_tokens": 776485388.0, "step": 20350 }, { "epoch": 2.5888563795954713, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.190042495727539, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8559754490852356, "num_tokens": 776529135.0, "step": 20351 }, { "epoch": 2.588983589874062, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.18579387664795, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8666731119155884, "num_tokens": 776564310.0, "step": 20352 }, { "epoch": 2.5891108001526524, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.167773246765137, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8868776559829712, "num_tokens": 776609004.0, "step": 20353 }, { "epoch": 2.589238010431243, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.224288940429688, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8806065320968628, "num_tokens": 776645796.0, "step": 20354 }, { "epoch": 2.5893652207098334, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.254717826843262, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8644338250160217, "num_tokens": 776684248.0, "step": 20355 }, { "epoch": 2.589492430988424, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.178977012634277, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8764539957046509, "num_tokens": 776722684.0, "step": 20356 }, { "epoch": 2.5896196412670145, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.226903915405273, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8777381777763367, "num_tokens": 776764635.0, "step": 20357 }, { "epoch": 2.589746851545605, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.176090240478516, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8633565902709961, "num_tokens": 776805683.0, "step": 20358 }, { "epoch": 2.5898740618241956, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.248827934265137, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8775795698165894, "num_tokens": 776841229.0, "step": 20359 }, { "epoch": 2.5900012721027856, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.14051628112793, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8814171552658081, "num_tokens": 776881319.0, "step": 20360 }, { "epoch": 2.5901284823813766, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.158570289611816, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8785531520843506, "num_tokens": 776919436.0, "step": 20361 }, { "epoch": 2.5902556926599667, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.250258445739746, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8562914729118347, "num_tokens": 776958644.0, "step": 20362 }, { "epoch": 2.5903829029385577, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.158302307128906, "learning_rate": 1e-06, "loss": 0.3465, "mean_token_accuracy": 0.8919038772583008, "num_tokens": 776994509.0, "step": 20363 }, { "epoch": 2.5905101132171477, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.15565013885498, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8796740770339966, "num_tokens": 777037557.0, "step": 20364 }, { "epoch": 2.5906373234957387, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.199414253234863, "learning_rate": 1e-06, "loss": 0.4662, "mean_token_accuracy": 0.8549191951751709, "num_tokens": 777077237.0, "step": 20365 }, { "epoch": 2.590764533774329, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.178808212280273, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8812088966369629, "num_tokens": 777120813.0, "step": 20366 }, { "epoch": 2.5908917440529193, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.161184310913086, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8613595962524414, "num_tokens": 777155809.0, "step": 20367 }, { "epoch": 2.59101895433151, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.212114334106445, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8746954202651978, "num_tokens": 777195868.0, "step": 20368 }, { "epoch": 2.5911461646101004, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.190073013305664, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8791454434394836, "num_tokens": 777233406.0, "step": 20369 }, { "epoch": 2.591273374888691, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.281085014343262, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8760757446289062, "num_tokens": 777272817.0, "step": 20370 }, { "epoch": 2.5914005851672814, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.20119571685791, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8704842329025269, "num_tokens": 777312398.0, "step": 20371 }, { "epoch": 2.591527795445872, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.196908950805664, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8839055299758911, "num_tokens": 777349375.0, "step": 20372 }, { "epoch": 2.5916550057244625, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.197738647460938, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8843832015991211, "num_tokens": 777394806.0, "step": 20373 }, { "epoch": 2.591782216003053, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.242072105407715, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8785490989685059, "num_tokens": 777434350.0, "step": 20374 }, { "epoch": 2.5919094262816436, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.174054145812988, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.848503589630127, "num_tokens": 777469690.0, "step": 20375 }, { "epoch": 2.592036636560234, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.232532501220703, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8687987923622131, "num_tokens": 777502095.0, "step": 20376 }, { "epoch": 2.5921638468388246, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.201251983642578, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8854618072509766, "num_tokens": 777540387.0, "step": 20377 }, { "epoch": 2.592291057117415, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.175000190734863, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8804901838302612, "num_tokens": 777581855.0, "step": 20378 }, { "epoch": 2.5924182673960057, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.221452713012695, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8750241994857788, "num_tokens": 777619852.0, "step": 20379 }, { "epoch": 2.592545477674596, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.162893295288086, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8682817220687866, "num_tokens": 777662772.0, "step": 20380 }, { "epoch": 2.5926726879531867, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.239748001098633, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8571929335594177, "num_tokens": 777701171.0, "step": 20381 }, { "epoch": 2.5927998982317773, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.198572158813477, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8611028790473938, "num_tokens": 777738245.0, "step": 20382 }, { "epoch": 2.592927108510368, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.226384162902832, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8743970394134521, "num_tokens": 777775732.0, "step": 20383 }, { "epoch": 2.5930543187889583, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.257378578186035, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8751346468925476, "num_tokens": 777811077.0, "step": 20384 }, { "epoch": 2.5931815290675484, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.17162799835205, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8720680475234985, "num_tokens": 777854107.0, "step": 20385 }, { "epoch": 2.5933087393461394, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.194726943969727, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8731930255889893, "num_tokens": 777891950.0, "step": 20386 }, { "epoch": 2.5934359496247295, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.17541217803955, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8917088508605957, "num_tokens": 777932638.0, "step": 20387 }, { "epoch": 2.5935631599033204, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.196089744567871, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8728898763656616, "num_tokens": 777977733.0, "step": 20388 }, { "epoch": 2.5936903701819105, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.25886058807373, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8789005279541016, "num_tokens": 778020451.0, "step": 20389 }, { "epoch": 2.5938175804605015, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.201306343078613, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8801959156990051, "num_tokens": 778059594.0, "step": 20390 }, { "epoch": 2.5939447907390916, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.15826416015625, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8778209686279297, "num_tokens": 778099235.0, "step": 20391 }, { "epoch": 2.594072001017682, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.23216724395752, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.88552325963974, "num_tokens": 778135430.0, "step": 20392 }, { "epoch": 2.5941992112962726, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.232851028442383, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8672664761543274, "num_tokens": 778175111.0, "step": 20393 }, { "epoch": 2.594326421574863, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.145607948303223, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8657513856887817, "num_tokens": 778211954.0, "step": 20394 }, { "epoch": 2.5944536318534537, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.270777702331543, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8591006994247437, "num_tokens": 778246580.0, "step": 20395 }, { "epoch": 2.594580842132044, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.19973087310791, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.880020797252655, "num_tokens": 778288659.0, "step": 20396 }, { "epoch": 2.5947080524106347, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.205718994140625, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8739875555038452, "num_tokens": 778325365.0, "step": 20397 }, { "epoch": 2.5948352626892253, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.236467361450195, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8805480003356934, "num_tokens": 778362982.0, "step": 20398 }, { "epoch": 2.594962472967816, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.224250793457031, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8608185648918152, "num_tokens": 778394230.0, "step": 20399 }, { "epoch": 2.5950896832464063, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.232277870178223, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.863694429397583, "num_tokens": 778434345.0, "step": 20400 }, { "epoch": 2.595216893524997, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.158037185668945, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8695135116577148, "num_tokens": 778478303.0, "step": 20401 }, { "epoch": 2.5953441038035874, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.215898513793945, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8799455761909485, "num_tokens": 778513915.0, "step": 20402 }, { "epoch": 2.595471314082178, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.169313430786133, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8897321224212646, "num_tokens": 778554294.0, "step": 20403 }, { "epoch": 2.5955985243607684, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.20627498626709, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8710758090019226, "num_tokens": 778592836.0, "step": 20404 }, { "epoch": 2.595725734639359, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.237630844116211, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8661547899246216, "num_tokens": 778629316.0, "step": 20405 }, { "epoch": 2.5958529449179495, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.26895523071289, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8704226016998291, "num_tokens": 778659065.0, "step": 20406 }, { "epoch": 2.59598015519654, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.219198226928711, "learning_rate": 1e-06, "loss": 0.3495, "mean_token_accuracy": 0.8894152641296387, "num_tokens": 778693949.0, "step": 20407 }, { "epoch": 2.5961073654751305, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.184161186218262, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.8794903755187988, "num_tokens": 778738045.0, "step": 20408 }, { "epoch": 2.596234575753721, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.205253601074219, "learning_rate": 1e-06, "loss": 0.5195, "mean_token_accuracy": 0.8324155807495117, "num_tokens": 778769578.0, "step": 20409 }, { "epoch": 2.596361786032311, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.17084789276123, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8699935078620911, "num_tokens": 778810052.0, "step": 20410 }, { "epoch": 2.596488996310902, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.163552284240723, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8671020865440369, "num_tokens": 778850868.0, "step": 20411 }, { "epoch": 2.596616206589492, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.160377502441406, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8801948428153992, "num_tokens": 778890129.0, "step": 20412 }, { "epoch": 2.596743416868083, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.20681095123291, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8784230947494507, "num_tokens": 778923181.0, "step": 20413 }, { "epoch": 2.5968706271466733, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.219293594360352, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8657433986663818, "num_tokens": 778963328.0, "step": 20414 }, { "epoch": 2.596997837425264, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.17995834350586, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8841968774795532, "num_tokens": 779003932.0, "step": 20415 }, { "epoch": 2.5971250477038543, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.236003875732422, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8635869026184082, "num_tokens": 779043913.0, "step": 20416 }, { "epoch": 2.597252257982445, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.196189880371094, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8713223934173584, "num_tokens": 779077126.0, "step": 20417 }, { "epoch": 2.5973794682610354, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.150975227355957, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8783562183380127, "num_tokens": 779111544.0, "step": 20418 }, { "epoch": 2.597506678539626, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.204277038574219, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8731223940849304, "num_tokens": 779145493.0, "step": 20419 }, { "epoch": 2.5976338888182164, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.190680503845215, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8661268353462219, "num_tokens": 779189216.0, "step": 20420 }, { "epoch": 2.597761099096807, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.235651016235352, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8811880350112915, "num_tokens": 779225914.0, "step": 20421 }, { "epoch": 2.5978883093753975, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.16950798034668, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8781021237373352, "num_tokens": 779264118.0, "step": 20422 }, { "epoch": 2.598015519653988, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.16689395904541, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8771294951438904, "num_tokens": 779301694.0, "step": 20423 }, { "epoch": 2.5981427299325786, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.164290428161621, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8731170892715454, "num_tokens": 779339012.0, "step": 20424 }, { "epoch": 2.598269940211169, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.16891098022461, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.878128170967102, "num_tokens": 779376358.0, "step": 20425 }, { "epoch": 2.5983971504897596, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.24081039428711, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8735947608947754, "num_tokens": 779416344.0, "step": 20426 }, { "epoch": 2.59852436076835, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.179072380065918, "learning_rate": 1e-06, "loss": 0.438, "mean_token_accuracy": 0.8626073598861694, "num_tokens": 779452189.0, "step": 20427 }, { "epoch": 2.5986515710469407, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.1561279296875, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8728306889533997, "num_tokens": 779493513.0, "step": 20428 }, { "epoch": 2.598778781325531, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.139274597167969, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8801762461662292, "num_tokens": 779533061.0, "step": 20429 }, { "epoch": 2.5989059916041217, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.266838073730469, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8670693635940552, "num_tokens": 779570159.0, "step": 20430 }, { "epoch": 2.5990332018827123, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.154948234558105, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8704897165298462, "num_tokens": 779609066.0, "step": 20431 }, { "epoch": 2.599160412161303, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.14705753326416, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.873779296875, "num_tokens": 779651087.0, "step": 20432 }, { "epoch": 2.5992876224398933, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.219930648803711, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8869074583053589, "num_tokens": 779689531.0, "step": 20433 }, { "epoch": 2.599414832718484, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.217240333557129, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8692018389701843, "num_tokens": 779724062.0, "step": 20434 }, { "epoch": 2.599542042997074, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.185562133789062, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8614163398742676, "num_tokens": 779766029.0, "step": 20435 }, { "epoch": 2.599669253275665, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.220575332641602, "learning_rate": 1e-06, "loss": 0.4476, "mean_token_accuracy": 0.8580113053321838, "num_tokens": 779811065.0, "step": 20436 }, { "epoch": 2.599796463554255, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.215662956237793, "learning_rate": 1e-06, "loss": 0.4423, "mean_token_accuracy": 0.8605085611343384, "num_tokens": 779852442.0, "step": 20437 }, { "epoch": 2.599923673832846, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.214838981628418, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8777278065681458, "num_tokens": 779885993.0, "step": 20438 }, { "epoch": 2.600050884111436, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.251996040344238, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8779538869857788, "num_tokens": 779927791.0, "step": 20439 }, { "epoch": 2.6001780943900266, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.203937530517578, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8774853348731995, "num_tokens": 779961517.0, "step": 20440 }, { "epoch": 2.600305304668617, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.233515739440918, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.8610472679138184, "num_tokens": 779992902.0, "step": 20441 }, { "epoch": 2.6004325149472076, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.219715118408203, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8741445541381836, "num_tokens": 780032849.0, "step": 20442 }, { "epoch": 2.600559725225798, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.100395202636719, "learning_rate": 1e-06, "loss": 0.3733, "mean_token_accuracy": 0.8857861757278442, "num_tokens": 780072563.0, "step": 20443 }, { "epoch": 2.6006869355043887, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.233867645263672, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8734536170959473, "num_tokens": 780115299.0, "step": 20444 }, { "epoch": 2.600814145782979, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.207204818725586, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8725497722625732, "num_tokens": 780155792.0, "step": 20445 }, { "epoch": 2.6009413560615697, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.188326835632324, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8693041801452637, "num_tokens": 780189040.0, "step": 20446 }, { "epoch": 2.6010685663401603, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.226053237915039, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8725754022598267, "num_tokens": 780229463.0, "step": 20447 }, { "epoch": 2.601195776618751, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.237784385681152, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8857666254043579, "num_tokens": 780265961.0, "step": 20448 }, { "epoch": 2.6013229868973413, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.154952049255371, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8686662316322327, "num_tokens": 780303220.0, "step": 20449 }, { "epoch": 2.601450197175932, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.272733688354492, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8700956106185913, "num_tokens": 780340456.0, "step": 20450 }, { "epoch": 2.6015774074545224, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.176346778869629, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8720479011535645, "num_tokens": 780381402.0, "step": 20451 }, { "epoch": 2.601704617733113, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.216851234436035, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8776699304580688, "num_tokens": 780422571.0, "step": 20452 }, { "epoch": 2.6018318280117034, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.171101570129395, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8698229193687439, "num_tokens": 780462447.0, "step": 20453 }, { "epoch": 2.601959038290294, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.224137306213379, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8735018968582153, "num_tokens": 780499183.0, "step": 20454 }, { "epoch": 2.6020862485688845, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.213597297668457, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8614802360534668, "num_tokens": 780539216.0, "step": 20455 }, { "epoch": 2.602213458847475, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.193753242492676, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8796391487121582, "num_tokens": 780578725.0, "step": 20456 }, { "epoch": 2.6023406691260655, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.278580665588379, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8743377923965454, "num_tokens": 780615175.0, "step": 20457 }, { "epoch": 2.6024678794046556, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.218162536621094, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8765316009521484, "num_tokens": 780651186.0, "step": 20458 }, { "epoch": 2.6025950896832466, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.237831115722656, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8663453459739685, "num_tokens": 780685550.0, "step": 20459 }, { "epoch": 2.6027222999618367, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.206522941589355, "learning_rate": 1e-06, "loss": 0.3324, "mean_token_accuracy": 0.8936101198196411, "num_tokens": 780725477.0, "step": 20460 }, { "epoch": 2.6028495102404277, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.193144798278809, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8760323524475098, "num_tokens": 780762481.0, "step": 20461 }, { "epoch": 2.6029767205190177, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.244730949401855, "learning_rate": 1e-06, "loss": 0.4631, "mean_token_accuracy": 0.8536849021911621, "num_tokens": 780800305.0, "step": 20462 }, { "epoch": 2.6031039307976087, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.20825481414795, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8755280375480652, "num_tokens": 780836271.0, "step": 20463 }, { "epoch": 2.603231141076199, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.13744068145752, "learning_rate": 1e-06, "loss": 0.3586, "mean_token_accuracy": 0.8847523927688599, "num_tokens": 780877363.0, "step": 20464 }, { "epoch": 2.6033583513547893, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.228246688842773, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8625611662864685, "num_tokens": 780922982.0, "step": 20465 }, { "epoch": 2.60348556163338, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.235718727111816, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8723167181015015, "num_tokens": 780957925.0, "step": 20466 }, { "epoch": 2.6036127719119704, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.19177532196045, "learning_rate": 1e-06, "loss": 0.47, "mean_token_accuracy": 0.8549141883850098, "num_tokens": 780996922.0, "step": 20467 }, { "epoch": 2.603739982190561, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.226353645324707, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8564872741699219, "num_tokens": 781037209.0, "step": 20468 }, { "epoch": 2.6038671924691514, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.213735580444336, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8665554523468018, "num_tokens": 781078077.0, "step": 20469 }, { "epoch": 2.603994402747742, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.231555938720703, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8722830414772034, "num_tokens": 781122516.0, "step": 20470 }, { "epoch": 2.6041216130263325, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.21568775177002, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8657691478729248, "num_tokens": 781156290.0, "step": 20471 }, { "epoch": 2.604248823304923, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.184707641601562, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8812286257743835, "num_tokens": 781192769.0, "step": 20472 }, { "epoch": 2.6043760335835135, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.22326946258545, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8774402141571045, "num_tokens": 781227594.0, "step": 20473 }, { "epoch": 2.604503243862104, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.22590160369873, "learning_rate": 1e-06, "loss": 0.3558, "mean_token_accuracy": 0.8850773572921753, "num_tokens": 781259369.0, "step": 20474 }, { "epoch": 2.6046304541406946, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25815486907959, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8726369738578796, "num_tokens": 781296788.0, "step": 20475 }, { "epoch": 2.604757664419285, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.14583969116211, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8754345178604126, "num_tokens": 781333103.0, "step": 20476 }, { "epoch": 2.6048848746978757, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.154026985168457, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8817974925041199, "num_tokens": 781372691.0, "step": 20477 }, { "epoch": 2.605012084976466, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.252250671386719, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8693849444389343, "num_tokens": 781410480.0, "step": 20478 }, { "epoch": 2.6051392952550567, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.185848236083984, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8639543056488037, "num_tokens": 781454110.0, "step": 20479 }, { "epoch": 2.6052665055336472, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.139893531799316, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8707579374313354, "num_tokens": 781494034.0, "step": 20480 }, { "epoch": 2.6053937158122378, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.220152854919434, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8706226348876953, "num_tokens": 781536225.0, "step": 20481 }, { "epoch": 2.6055209260908283, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.18431282043457, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8793697357177734, "num_tokens": 781573987.0, "step": 20482 }, { "epoch": 2.6056481363694184, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.222217559814453, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8782933950424194, "num_tokens": 781608987.0, "step": 20483 }, { "epoch": 2.6057753466480094, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.249811172485352, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8797516822814941, "num_tokens": 781646185.0, "step": 20484 }, { "epoch": 2.6059025569265994, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.207860946655273, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8640878796577454, "num_tokens": 781682999.0, "step": 20485 }, { "epoch": 2.6060297672051904, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.212510108947754, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8711060285568237, "num_tokens": 781726052.0, "step": 20486 }, { "epoch": 2.6061569774837805, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.25894832611084, "learning_rate": 1e-06, "loss": 0.3542, "mean_token_accuracy": 0.8867128491401672, "num_tokens": 781760562.0, "step": 20487 }, { "epoch": 2.6062841877623715, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.176493644714355, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8621656894683838, "num_tokens": 781803765.0, "step": 20488 }, { "epoch": 2.6064113980409616, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.226191520690918, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8748610019683838, "num_tokens": 781845419.0, "step": 20489 }, { "epoch": 2.606538608319552, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2106294631958, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8820555806159973, "num_tokens": 781883371.0, "step": 20490 }, { "epoch": 2.6066658185981426, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.232051849365234, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8720774054527283, "num_tokens": 781919669.0, "step": 20491 }, { "epoch": 2.606793028876733, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.183255195617676, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8799735307693481, "num_tokens": 781953212.0, "step": 20492 }, { "epoch": 2.6069202391553237, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.208988189697266, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8770492076873779, "num_tokens": 781996824.0, "step": 20493 }, { "epoch": 2.607047449433914, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.19793701171875, "learning_rate": 1e-06, "loss": 0.3363, "mean_token_accuracy": 0.8923797607421875, "num_tokens": 782035350.0, "step": 20494 }, { "epoch": 2.6071746597125047, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.151601791381836, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8684753179550171, "num_tokens": 782073217.0, "step": 20495 }, { "epoch": 2.6073018699910953, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.253068923950195, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8671254515647888, "num_tokens": 782110777.0, "step": 20496 }, { "epoch": 2.607429080269686, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.207046508789062, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8673852682113647, "num_tokens": 782153300.0, "step": 20497 }, { "epoch": 2.6075562905482763, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.151021957397461, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8786060810089111, "num_tokens": 782191945.0, "step": 20498 }, { "epoch": 2.607683500826867, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.25272274017334, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8835099339485168, "num_tokens": 782230472.0, "step": 20499 }, { "epoch": 2.6078107111054574, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.193594932556152, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.8651313781738281, "num_tokens": 782263868.0, "step": 20500 }, { "epoch": 2.607937921384048, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.16318130493164, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8757534027099609, "num_tokens": 782299781.0, "step": 20501 }, { "epoch": 2.6080651316626384, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.17076587677002, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8670310378074646, "num_tokens": 782343376.0, "step": 20502 }, { "epoch": 2.608192341941229, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.171039581298828, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8856083154678345, "num_tokens": 782389894.0, "step": 20503 }, { "epoch": 2.6083195522198195, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.15822696685791, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8773025274276733, "num_tokens": 782430055.0, "step": 20504 }, { "epoch": 2.60844676249841, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.215850830078125, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8751949667930603, "num_tokens": 782467211.0, "step": 20505 }, { "epoch": 2.6085739727770005, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.191279411315918, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.864055871963501, "num_tokens": 782505913.0, "step": 20506 }, { "epoch": 2.608701183055591, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.167529106140137, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.872613251209259, "num_tokens": 782541858.0, "step": 20507 }, { "epoch": 2.608828393334181, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.208309173583984, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8730671405792236, "num_tokens": 782575854.0, "step": 20508 }, { "epoch": 2.608955603612772, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.240181922912598, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8751109838485718, "num_tokens": 782614005.0, "step": 20509 }, { "epoch": 2.609082813891362, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.18469524383545, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.88096684217453, "num_tokens": 782650147.0, "step": 20510 }, { "epoch": 2.609210024169953, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.204087257385254, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8728886842727661, "num_tokens": 782683215.0, "step": 20511 }, { "epoch": 2.6093372344485433, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.212626457214355, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8795222043991089, "num_tokens": 782721184.0, "step": 20512 }, { "epoch": 2.609464444727134, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2130126953125, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8682571649551392, "num_tokens": 782759891.0, "step": 20513 }, { "epoch": 2.6095916550057243, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.233473777770996, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8631775379180908, "num_tokens": 782796398.0, "step": 20514 }, { "epoch": 2.609718865284315, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.15346908569336, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8831747770309448, "num_tokens": 782835968.0, "step": 20515 }, { "epoch": 2.6098460755629054, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.221307754516602, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8848034143447876, "num_tokens": 782871367.0, "step": 20516 }, { "epoch": 2.609973285841496, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.18601131439209, "learning_rate": 1e-06, "loss": 0.3458, "mean_token_accuracy": 0.8940987586975098, "num_tokens": 782905766.0, "step": 20517 }, { "epoch": 2.6101004961200864, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.231396675109863, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8776405453681946, "num_tokens": 782942256.0, "step": 20518 }, { "epoch": 2.610227706398677, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.20392894744873, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8832115530967712, "num_tokens": 782976156.0, "step": 20519 }, { "epoch": 2.6103549166772675, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.177092552185059, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8822883367538452, "num_tokens": 783011450.0, "step": 20520 }, { "epoch": 2.610482126955858, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.261730194091797, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8789048790931702, "num_tokens": 783050355.0, "step": 20521 }, { "epoch": 2.6106093372344485, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.249835014343262, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.884405255317688, "num_tokens": 783091621.0, "step": 20522 }, { "epoch": 2.610736547513039, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.168164253234863, "learning_rate": 1e-06, "loss": 0.4414, "mean_token_accuracy": 0.8631607294082642, "num_tokens": 783130471.0, "step": 20523 }, { "epoch": 2.6108637577916296, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.221495628356934, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8785771727561951, "num_tokens": 783168340.0, "step": 20524 }, { "epoch": 2.61099096807022, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.20779037475586, "learning_rate": 1e-06, "loss": 0.4515, "mean_token_accuracy": 0.8557443618774414, "num_tokens": 783207279.0, "step": 20525 }, { "epoch": 2.6111181783488107, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.26119613647461, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8672505617141724, "num_tokens": 783243010.0, "step": 20526 }, { "epoch": 2.611245388627401, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.167628288269043, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.863947331905365, "num_tokens": 783287548.0, "step": 20527 }, { "epoch": 2.6113725989059917, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.217175483703613, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8719215393066406, "num_tokens": 783331549.0, "step": 20528 }, { "epoch": 2.6114998091845822, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.201058387756348, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8834289312362671, "num_tokens": 783369445.0, "step": 20529 }, { "epoch": 2.6116270194631728, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2191162109375, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.869604229927063, "num_tokens": 783413218.0, "step": 20530 }, { "epoch": 2.6117542297417633, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.189525604248047, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8733930587768555, "num_tokens": 783449944.0, "step": 20531 }, { "epoch": 2.611881440020354, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.255572319030762, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8741953372955322, "num_tokens": 783486142.0, "step": 20532 }, { "epoch": 2.612008650298944, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.272629737854004, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8805059194564819, "num_tokens": 783521170.0, "step": 20533 }, { "epoch": 2.612135860577535, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.211689949035645, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8727561831474304, "num_tokens": 783563902.0, "step": 20534 }, { "epoch": 2.612263070856125, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.300488471984863, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8779831528663635, "num_tokens": 783597655.0, "step": 20535 }, { "epoch": 2.612390281134716, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.194682121276855, "learning_rate": 1e-06, "loss": 0.4276, "mean_token_accuracy": 0.8634113073348999, "num_tokens": 783635233.0, "step": 20536 }, { "epoch": 2.612517491413306, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.196537017822266, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8754854798316956, "num_tokens": 783679677.0, "step": 20537 }, { "epoch": 2.6126447016918966, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.207539558410645, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8797953128814697, "num_tokens": 783720608.0, "step": 20538 }, { "epoch": 2.612771911970487, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.251360893249512, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8840200901031494, "num_tokens": 783753198.0, "step": 20539 }, { "epoch": 2.6128991222490776, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.202754974365234, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8734657764434814, "num_tokens": 783791068.0, "step": 20540 }, { "epoch": 2.613026332527668, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.198826789855957, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8742169737815857, "num_tokens": 783831319.0, "step": 20541 }, { "epoch": 2.6131535428062587, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.269384384155273, "learning_rate": 1e-06, "loss": 0.3776, "mean_token_accuracy": 0.8783604502677917, "num_tokens": 783867537.0, "step": 20542 }, { "epoch": 2.613280753084849, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.169720649719238, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8690916895866394, "num_tokens": 783907787.0, "step": 20543 }, { "epoch": 2.6134079633634397, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.245579719543457, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.867608904838562, "num_tokens": 783948558.0, "step": 20544 }, { "epoch": 2.6135351736420303, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.18945026397705, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.8915313482284546, "num_tokens": 783986726.0, "step": 20545 }, { "epoch": 2.613662383920621, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.21689510345459, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8563356399536133, "num_tokens": 784023449.0, "step": 20546 }, { "epoch": 2.6137895941992113, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.237250328063965, "learning_rate": 1e-06, "loss": 0.3479, "mean_token_accuracy": 0.8908948302268982, "num_tokens": 784059529.0, "step": 20547 }, { "epoch": 2.613916804477802, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.214051246643066, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8746423721313477, "num_tokens": 784103530.0, "step": 20548 }, { "epoch": 2.6140440147563924, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.212799072265625, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8691036105155945, "num_tokens": 784141721.0, "step": 20549 }, { "epoch": 2.614171225034983, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.278406143188477, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8744707107543945, "num_tokens": 784175868.0, "step": 20550 }, { "epoch": 2.6142984353135734, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.198204040527344, "learning_rate": 1e-06, "loss": 0.3632, "mean_token_accuracy": 0.8865723609924316, "num_tokens": 784210646.0, "step": 20551 }, { "epoch": 2.614425645592164, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.245176315307617, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8794150352478027, "num_tokens": 784255588.0, "step": 20552 }, { "epoch": 2.6145528558707545, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.193953514099121, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8758352994918823, "num_tokens": 784292877.0, "step": 20553 }, { "epoch": 2.614680066149345, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.154656410217285, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8791729211807251, "num_tokens": 784336150.0, "step": 20554 }, { "epoch": 2.6148072764279355, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.251100540161133, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8863901495933533, "num_tokens": 784371547.0, "step": 20555 }, { "epoch": 2.6149344867065256, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.206400871276855, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.876400887966156, "num_tokens": 784411473.0, "step": 20556 }, { "epoch": 2.6150616969851166, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.251031875610352, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8654124736785889, "num_tokens": 784456713.0, "step": 20557 }, { "epoch": 2.6151889072637067, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.2178373336792, "learning_rate": 1e-06, "loss": 0.3532, "mean_token_accuracy": 0.8883167505264282, "num_tokens": 784493685.0, "step": 20558 }, { "epoch": 2.6153161175422976, "ewc_loss": 0.0380859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 14.213436126708984, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.885611891746521, "num_tokens": 784534641.0, "step": 20559 }, { "epoch": 2.6154433278208877, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.224642753601074, "learning_rate": 1e-06, "loss": 0.4723, "mean_token_accuracy": 0.8520023822784424, "num_tokens": 784576184.0, "step": 20560 }, { "epoch": 2.6155705380994787, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.224234580993652, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8793962001800537, "num_tokens": 784613851.0, "step": 20561 }, { "epoch": 2.615697748378069, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.182448387145996, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8765878677368164, "num_tokens": 784654133.0, "step": 20562 }, { "epoch": 2.6158249586566593, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.235913276672363, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8753467202186584, "num_tokens": 784690529.0, "step": 20563 }, { "epoch": 2.61595216893525, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.203462600708008, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8730558753013611, "num_tokens": 784731972.0, "step": 20564 }, { "epoch": 2.6160793792138404, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.184649467468262, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8704731464385986, "num_tokens": 784773275.0, "step": 20565 }, { "epoch": 2.616206589492431, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.228621482849121, "learning_rate": 1e-06, "loss": 0.4457, "mean_token_accuracy": 0.8600262403488159, "num_tokens": 784811468.0, "step": 20566 }, { "epoch": 2.6163337997710214, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.233901977539062, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.876976728439331, "num_tokens": 784850611.0, "step": 20567 }, { "epoch": 2.616461010049612, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.155622482299805, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8807076215744019, "num_tokens": 784886502.0, "step": 20568 }, { "epoch": 2.6165882203282025, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.26076889038086, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8667311668395996, "num_tokens": 784924489.0, "step": 20569 }, { "epoch": 2.616715430606793, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.28906536102295, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8802710771560669, "num_tokens": 784957224.0, "step": 20570 }, { "epoch": 2.6168426408853835, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.167880058288574, "learning_rate": 1e-06, "loss": 0.474, "mean_token_accuracy": 0.8503344058990479, "num_tokens": 784997754.0, "step": 20571 }, { "epoch": 2.616969851163974, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.171875953674316, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.860872209072113, "num_tokens": 785033932.0, "step": 20572 }, { "epoch": 2.6170970614425646, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.193102836608887, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8708226680755615, "num_tokens": 785073190.0, "step": 20573 }, { "epoch": 2.617224271721155, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.19772720336914, "learning_rate": 1e-06, "loss": 0.3452, "mean_token_accuracy": 0.890770673751831, "num_tokens": 785108397.0, "step": 20574 }, { "epoch": 2.6173514819997457, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.224955558776855, "learning_rate": 1e-06, "loss": 0.3249, "mean_token_accuracy": 0.900367259979248, "num_tokens": 785142671.0, "step": 20575 }, { "epoch": 2.617478692278336, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.170921325683594, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8730694055557251, "num_tokens": 785183467.0, "step": 20576 }, { "epoch": 2.6176059025569267, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.231649398803711, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8641472458839417, "num_tokens": 785222072.0, "step": 20577 }, { "epoch": 2.6177331128355172, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.187962532043457, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8634701371192932, "num_tokens": 785260458.0, "step": 20578 }, { "epoch": 2.6178603231141078, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.279729843139648, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8618803024291992, "num_tokens": 785302434.0, "step": 20579 }, { "epoch": 2.6179875333926983, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2811279296875, "learning_rate": 1e-06, "loss": 0.4429, "mean_token_accuracy": 0.8602555990219116, "num_tokens": 785338251.0, "step": 20580 }, { "epoch": 2.6181147436712884, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.154393196105957, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8705147504806519, "num_tokens": 785384789.0, "step": 20581 }, { "epoch": 2.6182419539498794, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.254552841186523, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8747990131378174, "num_tokens": 785424978.0, "step": 20582 }, { "epoch": 2.6183691642284694, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.212743759155273, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8662477731704712, "num_tokens": 785464421.0, "step": 20583 }, { "epoch": 2.6184963745070604, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.248140335083008, "learning_rate": 1e-06, "loss": 0.3343, "mean_token_accuracy": 0.8934416174888611, "num_tokens": 785496166.0, "step": 20584 }, { "epoch": 2.6186235847856505, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.217541694641113, "learning_rate": 1e-06, "loss": 0.4432, "mean_token_accuracy": 0.8602398633956909, "num_tokens": 785528504.0, "step": 20585 }, { "epoch": 2.6187507950642415, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.187698364257812, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8775124549865723, "num_tokens": 785570896.0, "step": 20586 }, { "epoch": 2.6188780053428315, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.160331726074219, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8709721565246582, "num_tokens": 785612464.0, "step": 20587 }, { "epoch": 2.619005215621422, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.24959659576416, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8735832571983337, "num_tokens": 785655500.0, "step": 20588 }, { "epoch": 2.6191324259000126, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.169541358947754, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8768365383148193, "num_tokens": 785691186.0, "step": 20589 }, { "epoch": 2.619259636178603, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.238184928894043, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.886253833770752, "num_tokens": 785728575.0, "step": 20590 }, { "epoch": 2.6193868464571937, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.17940902709961, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8719952702522278, "num_tokens": 785765782.0, "step": 20591 }, { "epoch": 2.619514056735784, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.18736457824707, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8786482810974121, "num_tokens": 785800130.0, "step": 20592 }, { "epoch": 2.6196412670143747, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.18407917022705, "learning_rate": 1e-06, "loss": 0.3136, "mean_token_accuracy": 0.9014382362365723, "num_tokens": 785836647.0, "step": 20593 }, { "epoch": 2.6197684772929652, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.20305061340332, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8689181804656982, "num_tokens": 785873897.0, "step": 20594 }, { "epoch": 2.6198956875715558, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.179886817932129, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8701212406158447, "num_tokens": 785914296.0, "step": 20595 }, { "epoch": 2.6200228978501463, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.188790321350098, "learning_rate": 1e-06, "loss": 0.4378, "mean_token_accuracy": 0.8622322082519531, "num_tokens": 785951426.0, "step": 20596 }, { "epoch": 2.620150108128737, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.22398853302002, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8761612772941589, "num_tokens": 785989906.0, "step": 20597 }, { "epoch": 2.6202773184073274, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.227056503295898, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8691151738166809, "num_tokens": 786024763.0, "step": 20598 }, { "epoch": 2.620404528685918, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.306615829467773, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8696298003196716, "num_tokens": 786063055.0, "step": 20599 }, { "epoch": 2.6205317389645084, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.326581001281738, "learning_rate": 1e-06, "loss": 0.4957, "mean_token_accuracy": 0.8492240309715271, "num_tokens": 786097675.0, "step": 20600 }, { "epoch": 2.620658949243099, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.281007766723633, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8849305510520935, "num_tokens": 786133721.0, "step": 20601 }, { "epoch": 2.6207861595216895, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.232786178588867, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8567382097244263, "num_tokens": 786179615.0, "step": 20602 }, { "epoch": 2.62091336980028, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.2467622756958, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8877031803131104, "num_tokens": 786214945.0, "step": 20603 }, { "epoch": 2.6210405800788705, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.274893760681152, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8755137920379639, "num_tokens": 786250880.0, "step": 20604 }, { "epoch": 2.621167790357461, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.284610748291016, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8733944296836853, "num_tokens": 786286095.0, "step": 20605 }, { "epoch": 2.621295000636051, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.254252433776855, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8787096738815308, "num_tokens": 786319547.0, "step": 20606 }, { "epoch": 2.621422210914642, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.213603973388672, "learning_rate": 1e-06, "loss": 0.3963, "mean_token_accuracy": 0.8739791512489319, "num_tokens": 786358699.0, "step": 20607 }, { "epoch": 2.621549421193232, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.248320579528809, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8779838681221008, "num_tokens": 786399732.0, "step": 20608 }, { "epoch": 2.621676631471823, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.20935344696045, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8659286499023438, "num_tokens": 786437375.0, "step": 20609 }, { "epoch": 2.6218038417504133, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.184082984924316, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8640672564506531, "num_tokens": 786475538.0, "step": 20610 }, { "epoch": 2.621931052029004, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.219351768493652, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8758702278137207, "num_tokens": 786516802.0, "step": 20611 }, { "epoch": 2.6220582623075943, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.223250389099121, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8745697736740112, "num_tokens": 786558470.0, "step": 20612 }, { "epoch": 2.622185472586185, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.15609359741211, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8686953186988831, "num_tokens": 786601163.0, "step": 20613 }, { "epoch": 2.6223126828647754, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.274328231811523, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.881242036819458, "num_tokens": 786636741.0, "step": 20614 }, { "epoch": 2.622439893143366, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.181017875671387, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8626677989959717, "num_tokens": 786672931.0, "step": 20615 }, { "epoch": 2.6225671034219564, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.220012664794922, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8867179751396179, "num_tokens": 786709310.0, "step": 20616 }, { "epoch": 2.622694313700547, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.269329071044922, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8797401189804077, "num_tokens": 786747113.0, "step": 20617 }, { "epoch": 2.6228215239791375, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25464916229248, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8729057312011719, "num_tokens": 786782033.0, "step": 20618 }, { "epoch": 2.622948734257728, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.213953971862793, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8750582933425903, "num_tokens": 786827229.0, "step": 20619 }, { "epoch": 2.6230759445363185, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.257880210876465, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8812853097915649, "num_tokens": 786861722.0, "step": 20620 }, { "epoch": 2.623203154814909, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.270707130432129, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8777533769607544, "num_tokens": 786896202.0, "step": 20621 }, { "epoch": 2.6233303650934996, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.248332023620605, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8759924173355103, "num_tokens": 786932154.0, "step": 20622 }, { "epoch": 2.62345757537209, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.202184677124023, "learning_rate": 1e-06, "loss": 0.4284, "mean_token_accuracy": 0.8654677867889404, "num_tokens": 786972184.0, "step": 20623 }, { "epoch": 2.6235847856506807, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.281859397888184, "learning_rate": 1e-06, "loss": 0.4221, "mean_token_accuracy": 0.8681658506393433, "num_tokens": 787015514.0, "step": 20624 }, { "epoch": 2.623711995929271, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.163286209106445, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8749867677688599, "num_tokens": 787054023.0, "step": 20625 }, { "epoch": 2.6238392062078617, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.29821491241455, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8710614442825317, "num_tokens": 787088858.0, "step": 20626 }, { "epoch": 2.6239664164864522, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.226491928100586, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8764240741729736, "num_tokens": 787124805.0, "step": 20627 }, { "epoch": 2.6240936267650428, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.229538917541504, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.886467456817627, "num_tokens": 787164922.0, "step": 20628 }, { "epoch": 2.6242208370436333, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.292616844177246, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8863265514373779, "num_tokens": 787204936.0, "step": 20629 }, { "epoch": 2.624348047322224, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.230817794799805, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8858687281608582, "num_tokens": 787238292.0, "step": 20630 }, { "epoch": 2.624475257600814, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.280430793762207, "learning_rate": 1e-06, "loss": 0.337, "mean_token_accuracy": 0.8933894634246826, "num_tokens": 787269054.0, "step": 20631 }, { "epoch": 2.624602467879405, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.265863418579102, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8729557991027832, "num_tokens": 787310022.0, "step": 20632 }, { "epoch": 2.624729678157995, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.165708541870117, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8639645576477051, "num_tokens": 787350917.0, "step": 20633 }, { "epoch": 2.624856888436586, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.203580856323242, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8830314874649048, "num_tokens": 787388187.0, "step": 20634 }, { "epoch": 2.624984098715176, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.242493629455566, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8809046745300293, "num_tokens": 787424998.0, "step": 20635 }, { "epoch": 2.6251113089937665, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.221165657043457, "learning_rate": 1e-06, "loss": 0.3601, "mean_token_accuracy": 0.8861596584320068, "num_tokens": 787465456.0, "step": 20636 }, { "epoch": 2.625238519272357, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.23405647277832, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8735455274581909, "num_tokens": 787504797.0, "step": 20637 }, { "epoch": 2.6253657295509476, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.196784973144531, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8839672803878784, "num_tokens": 787542911.0, "step": 20638 }, { "epoch": 2.625492939829538, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.232819557189941, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8656860589981079, "num_tokens": 787576143.0, "step": 20639 }, { "epoch": 2.6256201501081287, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.25195598602295, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8765437006950378, "num_tokens": 787619912.0, "step": 20640 }, { "epoch": 2.625747360386719, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.238616943359375, "learning_rate": 1e-06, "loss": 0.4213, "mean_token_accuracy": 0.8648837804794312, "num_tokens": 787658181.0, "step": 20641 }, { "epoch": 2.6258745706653097, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.254863739013672, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8657708168029785, "num_tokens": 787694889.0, "step": 20642 }, { "epoch": 2.6260017809439002, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.211891174316406, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.887866735458374, "num_tokens": 787737633.0, "step": 20643 }, { "epoch": 2.6261289912224908, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.2587890625, "learning_rate": 1e-06, "loss": 0.4422, "mean_token_accuracy": 0.8625777959823608, "num_tokens": 787779069.0, "step": 20644 }, { "epoch": 2.6262562015010813, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.186647415161133, "learning_rate": 1e-06, "loss": 0.4486, "mean_token_accuracy": 0.8621129989624023, "num_tokens": 787815544.0, "step": 20645 }, { "epoch": 2.626383411779672, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.269952774047852, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8692055344581604, "num_tokens": 787852684.0, "step": 20646 }, { "epoch": 2.6265106220582624, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.214877128601074, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8707274794578552, "num_tokens": 787893452.0, "step": 20647 }, { "epoch": 2.626637832336853, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.271036148071289, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8735126256942749, "num_tokens": 787929430.0, "step": 20648 }, { "epoch": 2.6267650426154434, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.234343528747559, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8720459938049316, "num_tokens": 787969768.0, "step": 20649 }, { "epoch": 2.626892252894034, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.229302406311035, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8738182783126831, "num_tokens": 788009605.0, "step": 20650 }, { "epoch": 2.6270194631726245, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.265495300292969, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8864946365356445, "num_tokens": 788044652.0, "step": 20651 }, { "epoch": 2.627146673451215, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.278799057006836, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8485301733016968, "num_tokens": 788085981.0, "step": 20652 }, { "epoch": 2.6272738837298055, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.301192283630371, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8717823624610901, "num_tokens": 788132196.0, "step": 20653 }, { "epoch": 2.6274010940083956, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.240453720092773, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8863843679428101, "num_tokens": 788160099.0, "step": 20654 }, { "epoch": 2.6275283042869866, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.274672508239746, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8731715083122253, "num_tokens": 788197264.0, "step": 20655 }, { "epoch": 2.6276555145655767, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.213294982910156, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8855680227279663, "num_tokens": 788232900.0, "step": 20656 }, { "epoch": 2.6277827248441676, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.270071983337402, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8689672946929932, "num_tokens": 788273720.0, "step": 20657 }, { "epoch": 2.6279099351227577, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.242547035217285, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8770332932472229, "num_tokens": 788308425.0, "step": 20658 }, { "epoch": 2.6280371454013487, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.248908042907715, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8663901090621948, "num_tokens": 788343701.0, "step": 20659 }, { "epoch": 2.628164355679939, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.269936561584473, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8592032194137573, "num_tokens": 788372789.0, "step": 20660 }, { "epoch": 2.6282915659585293, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.244129180908203, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8784688711166382, "num_tokens": 788408422.0, "step": 20661 }, { "epoch": 2.62841877623712, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.248146057128906, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8780015707015991, "num_tokens": 788446113.0, "step": 20662 }, { "epoch": 2.6285459865157104, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.266992568969727, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8683277368545532, "num_tokens": 788485009.0, "step": 20663 }, { "epoch": 2.628673196794301, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.246576309204102, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8771730661392212, "num_tokens": 788523446.0, "step": 20664 }, { "epoch": 2.6288004070728914, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.2689790725708, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.882342517375946, "num_tokens": 788557423.0, "step": 20665 }, { "epoch": 2.628927617351482, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.254785537719727, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8723885416984558, "num_tokens": 788588177.0, "step": 20666 }, { "epoch": 2.6290548276300725, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.309073448181152, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8752874732017517, "num_tokens": 788624877.0, "step": 20667 }, { "epoch": 2.629182037908663, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.257182121276855, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8839114308357239, "num_tokens": 788662532.0, "step": 20668 }, { "epoch": 2.6293092481872535, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.289359092712402, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8747256398200989, "num_tokens": 788702734.0, "step": 20669 }, { "epoch": 2.629436458465844, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.296835899353027, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.875117301940918, "num_tokens": 788738041.0, "step": 20670 }, { "epoch": 2.6295636687444346, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.352492332458496, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8757593631744385, "num_tokens": 788772998.0, "step": 20671 }, { "epoch": 2.629690879023025, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.204484939575195, "learning_rate": 1e-06, "loss": 0.4438, "mean_token_accuracy": 0.8589396476745605, "num_tokens": 788814310.0, "step": 20672 }, { "epoch": 2.6298180893016156, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.215517044067383, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8783628344535828, "num_tokens": 788852811.0, "step": 20673 }, { "epoch": 2.629945299580206, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.270005226135254, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8817734122276306, "num_tokens": 788888756.0, "step": 20674 }, { "epoch": 2.6300725098587967, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.261614799499512, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8775765895843506, "num_tokens": 788926292.0, "step": 20675 }, { "epoch": 2.6301997201373872, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.239055633544922, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8847495317459106, "num_tokens": 788965625.0, "step": 20676 }, { "epoch": 2.6303269304159778, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.226344108581543, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8784931302070618, "num_tokens": 789007390.0, "step": 20677 }, { "epoch": 2.6304541406945683, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.243597984313965, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8776425123214722, "num_tokens": 789047940.0, "step": 20678 }, { "epoch": 2.6305813509731584, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.260649681091309, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8745703101158142, "num_tokens": 789086321.0, "step": 20679 }, { "epoch": 2.6307085612517493, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.277222633361816, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8858392238616943, "num_tokens": 789121757.0, "step": 20680 }, { "epoch": 2.6308357715303394, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.256914138793945, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8885818719863892, "num_tokens": 789160076.0, "step": 20681 }, { "epoch": 2.6309629818089304, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.316317558288574, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8651549220085144, "num_tokens": 789197080.0, "step": 20682 }, { "epoch": 2.6310901920875205, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.245530128479004, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8688058853149414, "num_tokens": 789249876.0, "step": 20683 }, { "epoch": 2.6312174023661115, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.278290748596191, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8725515604019165, "num_tokens": 789290146.0, "step": 20684 }, { "epoch": 2.6313446126447015, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.269018173217773, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8778102397918701, "num_tokens": 789335610.0, "step": 20685 }, { "epoch": 2.631471822923292, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.25197696685791, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8699452877044678, "num_tokens": 789371956.0, "step": 20686 }, { "epoch": 2.6315990332018826, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.21573543548584, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8703811764717102, "num_tokens": 789415257.0, "step": 20687 }, { "epoch": 2.631726243480473, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.258667945861816, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8747385740280151, "num_tokens": 789452533.0, "step": 20688 }, { "epoch": 2.6318534537590637, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.286338806152344, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8721370697021484, "num_tokens": 789488780.0, "step": 20689 }, { "epoch": 2.631980664037654, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.21293830871582, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8714863061904907, "num_tokens": 789528169.0, "step": 20690 }, { "epoch": 2.6321078743162447, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.28005313873291, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8669328093528748, "num_tokens": 789568382.0, "step": 20691 }, { "epoch": 2.6322350845948352, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.234557151794434, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8716610670089722, "num_tokens": 789606895.0, "step": 20692 }, { "epoch": 2.6323622948734258, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.211620330810547, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8593384623527527, "num_tokens": 789649106.0, "step": 20693 }, { "epoch": 2.6324895051520163, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.224467277526855, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8783859610557556, "num_tokens": 789689449.0, "step": 20694 }, { "epoch": 2.632616715430607, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.260229110717773, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8847473859786987, "num_tokens": 789726179.0, "step": 20695 }, { "epoch": 2.6327439257091974, "ewc_loss": 0.038330078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.263691902160645, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8569247126579285, "num_tokens": 789764858.0, "step": 20696 }, { "epoch": 2.632871135987788, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.297125816345215, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.861666738986969, "num_tokens": 789799196.0, "step": 20697 }, { "epoch": 2.6329983462663784, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.228501319885254, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8720173835754395, "num_tokens": 789840498.0, "step": 20698 }, { "epoch": 2.633125556544969, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.172673225402832, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8778737783432007, "num_tokens": 789874269.0, "step": 20699 }, { "epoch": 2.6332527668235595, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.30783748626709, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8754822015762329, "num_tokens": 789914938.0, "step": 20700 }, { "epoch": 2.63337997710215, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.24787425994873, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8743933439254761, "num_tokens": 789959219.0, "step": 20701 }, { "epoch": 2.6335071873807405, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.29616928100586, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.867342472076416, "num_tokens": 789993707.0, "step": 20702 }, { "epoch": 2.633634397659331, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.222675323486328, "learning_rate": 1e-06, "loss": 0.4578, "mean_token_accuracy": 0.8543418645858765, "num_tokens": 790034285.0, "step": 20703 }, { "epoch": 2.633761607937921, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.277393341064453, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8771735429763794, "num_tokens": 790069714.0, "step": 20704 }, { "epoch": 2.633888818216512, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.261725425720215, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.872525691986084, "num_tokens": 790110733.0, "step": 20705 }, { "epoch": 2.634016028495102, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.293898582458496, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8786754012107849, "num_tokens": 790144468.0, "step": 20706 }, { "epoch": 2.634143238773693, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.21867561340332, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.871909499168396, "num_tokens": 790186881.0, "step": 20707 }, { "epoch": 2.6342704490522832, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.258563041687012, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8809517025947571, "num_tokens": 790225083.0, "step": 20708 }, { "epoch": 2.6343976593308738, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.274059295654297, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8737710118293762, "num_tokens": 790266822.0, "step": 20709 }, { "epoch": 2.6345248696094643, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.283965110778809, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.871036171913147, "num_tokens": 790305948.0, "step": 20710 }, { "epoch": 2.634652079888055, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.261263847351074, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8765867948532104, "num_tokens": 790341639.0, "step": 20711 }, { "epoch": 2.6347792901666454, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.267333984375, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8827771544456482, "num_tokens": 790377442.0, "step": 20712 }, { "epoch": 2.634906500445236, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.23599910736084, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8739526867866516, "num_tokens": 790414911.0, "step": 20713 }, { "epoch": 2.6350337107238264, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.362517356872559, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8724076151847839, "num_tokens": 790453968.0, "step": 20714 }, { "epoch": 2.635160921002417, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2489013671875, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8695447444915771, "num_tokens": 790493216.0, "step": 20715 }, { "epoch": 2.6352881312810075, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.309285163879395, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8695371747016907, "num_tokens": 790529380.0, "step": 20716 }, { "epoch": 2.635415341559598, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.239422798156738, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8688351511955261, "num_tokens": 790569568.0, "step": 20717 }, { "epoch": 2.6355425518381885, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.370524406433105, "learning_rate": 1e-06, "loss": 0.3709, "mean_token_accuracy": 0.8801988363265991, "num_tokens": 790602269.0, "step": 20718 }, { "epoch": 2.635669762116779, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.259373664855957, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8628158569335938, "num_tokens": 790639723.0, "step": 20719 }, { "epoch": 2.6357969723953696, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.2904691696167, "learning_rate": 1e-06, "loss": 0.4467, "mean_token_accuracy": 0.8609142303466797, "num_tokens": 790676103.0, "step": 20720 }, { "epoch": 2.63592418267396, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.280218124389648, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8760955929756165, "num_tokens": 790712118.0, "step": 20721 }, { "epoch": 2.6360513929525506, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.32616901397705, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.873496413230896, "num_tokens": 790747079.0, "step": 20722 }, { "epoch": 2.636178603231141, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.297232627868652, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8638728260993958, "num_tokens": 790781741.0, "step": 20723 }, { "epoch": 2.6363058135097317, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.264747619628906, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8724509477615356, "num_tokens": 790820243.0, "step": 20724 }, { "epoch": 2.6364330237883222, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.259827613830566, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.889043927192688, "num_tokens": 790851492.0, "step": 20725 }, { "epoch": 2.6365602340669128, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.290367126464844, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8790228366851807, "num_tokens": 790882293.0, "step": 20726 }, { "epoch": 2.636687444345503, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.306242942810059, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8700780272483826, "num_tokens": 790922188.0, "step": 20727 }, { "epoch": 2.636814654624094, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.229510307312012, "learning_rate": 1e-06, "loss": 0.3645, "mean_token_accuracy": 0.883813202381134, "num_tokens": 790960675.0, "step": 20728 }, { "epoch": 2.636941864902684, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.263745307922363, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8823915719985962, "num_tokens": 790999352.0, "step": 20729 }, { "epoch": 2.637069075181275, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.288216590881348, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8750184774398804, "num_tokens": 791035898.0, "step": 20730 }, { "epoch": 2.637196285459865, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.284786224365234, "learning_rate": 1e-06, "loss": 0.4703, "mean_token_accuracy": 0.8500463962554932, "num_tokens": 791079731.0, "step": 20731 }, { "epoch": 2.637323495738456, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.274739265441895, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8860488533973694, "num_tokens": 791118565.0, "step": 20732 }, { "epoch": 2.637450706017046, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.303704261779785, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8753421306610107, "num_tokens": 791154876.0, "step": 20733 }, { "epoch": 2.6375779162956365, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.316006660461426, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8740573525428772, "num_tokens": 791196003.0, "step": 20734 }, { "epoch": 2.637705126574227, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.28000545501709, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8725582361221313, "num_tokens": 791236462.0, "step": 20735 }, { "epoch": 2.6378323368528176, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.297733306884766, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8601944446563721, "num_tokens": 791271918.0, "step": 20736 }, { "epoch": 2.637959547131408, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.224348068237305, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8855977654457092, "num_tokens": 791309813.0, "step": 20737 }, { "epoch": 2.6380867574099987, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.315858840942383, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8828710317611694, "num_tokens": 791338854.0, "step": 20738 }, { "epoch": 2.638213967688589, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.276717185974121, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8786337375640869, "num_tokens": 791376580.0, "step": 20739 }, { "epoch": 2.6383411779671797, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.319379806518555, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8694326877593994, "num_tokens": 791415931.0, "step": 20740 }, { "epoch": 2.6384683882457702, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.266523361206055, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8704210519790649, "num_tokens": 791448201.0, "step": 20741 }, { "epoch": 2.6385955985243608, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.266830444335938, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8670288324356079, "num_tokens": 791489183.0, "step": 20742 }, { "epoch": 2.6387228088029513, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.274977684020996, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8749432563781738, "num_tokens": 791528414.0, "step": 20743 }, { "epoch": 2.638850019081542, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25731086730957, "learning_rate": 1e-06, "loss": 0.3417, "mean_token_accuracy": 0.8928967714309692, "num_tokens": 791566379.0, "step": 20744 }, { "epoch": 2.6389772293601323, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.307198524475098, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8772070407867432, "num_tokens": 791599789.0, "step": 20745 }, { "epoch": 2.639104439638723, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.208280563354492, "learning_rate": 1e-06, "loss": 0.4408, "mean_token_accuracy": 0.8615915775299072, "num_tokens": 791642333.0, "step": 20746 }, { "epoch": 2.6392316499173134, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.258399963378906, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8747638463973999, "num_tokens": 791687190.0, "step": 20747 }, { "epoch": 2.639358860195904, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.208386421203613, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8588645458221436, "num_tokens": 791722806.0, "step": 20748 }, { "epoch": 2.6394860704744945, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.325597763061523, "learning_rate": 1e-06, "loss": 0.4454, "mean_token_accuracy": 0.8594002723693848, "num_tokens": 791757726.0, "step": 20749 }, { "epoch": 2.639613280753085, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.221982955932617, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8811658024787903, "num_tokens": 791795160.0, "step": 20750 }, { "epoch": 2.6397404910316755, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.259942054748535, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.885033905506134, "num_tokens": 791831682.0, "step": 20751 }, { "epoch": 2.6398677013102656, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.211135864257812, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8828186392784119, "num_tokens": 791863230.0, "step": 20752 }, { "epoch": 2.6399949115888566, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.295755386352539, "learning_rate": 1e-06, "loss": 0.3525, "mean_token_accuracy": 0.886718213558197, "num_tokens": 791904323.0, "step": 20753 }, { "epoch": 2.6401221218674467, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.242871284484863, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8670471906661987, "num_tokens": 791951836.0, "step": 20754 }, { "epoch": 2.6402493321460376, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.275835037231445, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8761790990829468, "num_tokens": 791989782.0, "step": 20755 }, { "epoch": 2.6403765424246277, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.311964988708496, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8752566576004028, "num_tokens": 792024038.0, "step": 20756 }, { "epoch": 2.6405037527032187, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.227166175842285, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.874086856842041, "num_tokens": 792067893.0, "step": 20757 }, { "epoch": 2.6406309629818088, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.224318504333496, "learning_rate": 1e-06, "loss": 0.3628, "mean_token_accuracy": 0.8833004236221313, "num_tokens": 792108593.0, "step": 20758 }, { "epoch": 2.6407581732603993, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.272957801818848, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8865160942077637, "num_tokens": 792141201.0, "step": 20759 }, { "epoch": 2.64088538353899, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.23403263092041, "learning_rate": 1e-06, "loss": 0.3425, "mean_token_accuracy": 0.8936690092086792, "num_tokens": 792182908.0, "step": 20760 }, { "epoch": 2.6410125938175804, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.221208572387695, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8749167323112488, "num_tokens": 792225223.0, "step": 20761 }, { "epoch": 2.641139804096171, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.267735481262207, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8571594953536987, "num_tokens": 792258131.0, "step": 20762 }, { "epoch": 2.6412670143747614, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.237652778625488, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8803829550743103, "num_tokens": 792291429.0, "step": 20763 }, { "epoch": 2.641394224653352, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.276457786560059, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8703560829162598, "num_tokens": 792327777.0, "step": 20764 }, { "epoch": 2.6415214349319425, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.256897926330566, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8767534494400024, "num_tokens": 792367918.0, "step": 20765 }, { "epoch": 2.641648645210533, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.279585838317871, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.878149151802063, "num_tokens": 792405550.0, "step": 20766 }, { "epoch": 2.6417758554891235, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.295204162597656, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8766453266143799, "num_tokens": 792440948.0, "step": 20767 }, { "epoch": 2.641903065767714, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.216886520385742, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8819714784622192, "num_tokens": 792488065.0, "step": 20768 }, { "epoch": 2.6420302760463046, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.287019729614258, "learning_rate": 1e-06, "loss": 0.3678, "mean_token_accuracy": 0.8820348381996155, "num_tokens": 792522845.0, "step": 20769 }, { "epoch": 2.642157486324895, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.20750904083252, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8849829435348511, "num_tokens": 792561118.0, "step": 20770 }, { "epoch": 2.6422846966034856, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.303082466125488, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8737261295318604, "num_tokens": 792600125.0, "step": 20771 }, { "epoch": 2.642411906882076, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.261329650878906, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.880346417427063, "num_tokens": 792633856.0, "step": 20772 }, { "epoch": 2.6425391171606667, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.33510971069336, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.871198296546936, "num_tokens": 792672427.0, "step": 20773 }, { "epoch": 2.6426663274392572, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.210847854614258, "learning_rate": 1e-06, "loss": 0.4485, "mean_token_accuracy": 0.852800726890564, "num_tokens": 792713856.0, "step": 20774 }, { "epoch": 2.6427935377178478, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.289149284362793, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8670486807823181, "num_tokens": 792757602.0, "step": 20775 }, { "epoch": 2.6429207479964383, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.293184280395508, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8661543130874634, "num_tokens": 792797067.0, "step": 20776 }, { "epoch": 2.6430479582750284, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.254637718200684, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8779324293136597, "num_tokens": 792835748.0, "step": 20777 }, { "epoch": 2.6431751685536193, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.302149772644043, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8760181665420532, "num_tokens": 792875039.0, "step": 20778 }, { "epoch": 2.6433023788322094, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.286466598510742, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.879888653755188, "num_tokens": 792916580.0, "step": 20779 }, { "epoch": 2.6434295891108004, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.25513744354248, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8765205144882202, "num_tokens": 792952507.0, "step": 20780 }, { "epoch": 2.6435567993893905, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.291764259338379, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8779172897338867, "num_tokens": 792994046.0, "step": 20781 }, { "epoch": 2.6436840096679814, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.284344673156738, "learning_rate": 1e-06, "loss": 0.3463, "mean_token_accuracy": 0.8930284976959229, "num_tokens": 793033855.0, "step": 20782 }, { "epoch": 2.6438112199465715, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.281377792358398, "learning_rate": 1e-06, "loss": 0.3779, "mean_token_accuracy": 0.8791399598121643, "num_tokens": 793072226.0, "step": 20783 }, { "epoch": 2.643938430225162, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.306143760681152, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8785136342048645, "num_tokens": 793110884.0, "step": 20784 }, { "epoch": 2.6440656405037526, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.30172061920166, "learning_rate": 1e-06, "loss": 0.4603, "mean_token_accuracy": 0.8569711446762085, "num_tokens": 793151885.0, "step": 20785 }, { "epoch": 2.644192850782343, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.318674087524414, "learning_rate": 1e-06, "loss": 0.4601, "mean_token_accuracy": 0.8527964949607849, "num_tokens": 793187253.0, "step": 20786 }, { "epoch": 2.6443200610609336, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.32715129852295, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8797534704208374, "num_tokens": 793231249.0, "step": 20787 }, { "epoch": 2.644447271339524, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.281818389892578, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.88617342710495, "num_tokens": 793269918.0, "step": 20788 }, { "epoch": 2.6445744816181147, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.23801326751709, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8845393657684326, "num_tokens": 793306856.0, "step": 20789 }, { "epoch": 2.6447016918967052, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.326275825500488, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8661040663719177, "num_tokens": 793347866.0, "step": 20790 }, { "epoch": 2.6448289021752958, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.232476234436035, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8731422424316406, "num_tokens": 793391037.0, "step": 20791 }, { "epoch": 2.6449561124538863, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.334538459777832, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8618048429489136, "num_tokens": 793423560.0, "step": 20792 }, { "epoch": 2.645083322732477, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.271553993225098, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8818353414535522, "num_tokens": 793460215.0, "step": 20793 }, { "epoch": 2.6452105330110673, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.289193153381348, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8634654879570007, "num_tokens": 793492671.0, "step": 20794 }, { "epoch": 2.645337743289658, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.25655460357666, "learning_rate": 1e-06, "loss": 0.3948, "mean_token_accuracy": 0.8734238147735596, "num_tokens": 793529526.0, "step": 20795 }, { "epoch": 2.6454649535682484, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.281373977661133, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.8828747272491455, "num_tokens": 793566748.0, "step": 20796 }, { "epoch": 2.645592163846839, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.269367218017578, "learning_rate": 1e-06, "loss": 0.3504, "mean_token_accuracy": 0.8893502950668335, "num_tokens": 793600703.0, "step": 20797 }, { "epoch": 2.6457193741254295, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.340886116027832, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8745976686477661, "num_tokens": 793639226.0, "step": 20798 }, { "epoch": 2.64584658440402, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.29675006866455, "learning_rate": 1e-06, "loss": 0.35, "mean_token_accuracy": 0.8874829411506653, "num_tokens": 793676147.0, "step": 20799 }, { "epoch": 2.6459737946826105, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 14.311595916748047, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8843927979469299, "num_tokens": 793715778.0, "step": 20800 }, { "epoch": 2.646101004961201, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.31115436553955, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.889968752861023, "num_tokens": 793752096.0, "step": 20801 }, { "epoch": 2.646228215239791, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.289653778076172, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8698574900627136, "num_tokens": 793789809.0, "step": 20802 }, { "epoch": 2.646355425518382, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.318381309509277, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8710798025131226, "num_tokens": 793826505.0, "step": 20803 }, { "epoch": 2.646482635796972, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.28360652923584, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8739538192749023, "num_tokens": 793868360.0, "step": 20804 }, { "epoch": 2.646609846075563, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.296600341796875, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8767162561416626, "num_tokens": 793906566.0, "step": 20805 }, { "epoch": 2.6467370563541532, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.266878128051758, "learning_rate": 1e-06, "loss": 0.413, "mean_token_accuracy": 0.8727911114692688, "num_tokens": 793947814.0, "step": 20806 }, { "epoch": 2.6468642666327438, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.318739891052246, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8804219961166382, "num_tokens": 793987391.0, "step": 20807 }, { "epoch": 2.6469914769113343, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.217903137207031, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8855963945388794, "num_tokens": 794024045.0, "step": 20808 }, { "epoch": 2.647118687189925, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.308701515197754, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8791567087173462, "num_tokens": 794063405.0, "step": 20809 }, { "epoch": 2.6472458974685154, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.323616027832031, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8515056371688843, "num_tokens": 794099729.0, "step": 20810 }, { "epoch": 2.647373107747106, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.267300605773926, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8676249980926514, "num_tokens": 794139669.0, "step": 20811 }, { "epoch": 2.6475003180256964, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.292081832885742, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8763009905815125, "num_tokens": 794176906.0, "step": 20812 }, { "epoch": 2.647627528304287, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.281251907348633, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.877374529838562, "num_tokens": 794216594.0, "step": 20813 }, { "epoch": 2.6477547385828775, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.295456886291504, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8777881860733032, "num_tokens": 794248379.0, "step": 20814 }, { "epoch": 2.647881948861468, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.22979736328125, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8523590564727783, "num_tokens": 794293604.0, "step": 20815 }, { "epoch": 2.6480091591400585, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.348328590393066, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8651432394981384, "num_tokens": 794332084.0, "step": 20816 }, { "epoch": 2.648136369418649, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.256839752197266, "learning_rate": 1e-06, "loss": 0.3503, "mean_token_accuracy": 0.8899173140525818, "num_tokens": 794369158.0, "step": 20817 }, { "epoch": 2.6482635796972396, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.277899742126465, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8782019019126892, "num_tokens": 794410531.0, "step": 20818 }, { "epoch": 2.64839078997583, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.266907691955566, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8755280375480652, "num_tokens": 794457556.0, "step": 20819 }, { "epoch": 2.6485180002544206, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.32900333404541, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8637774586677551, "num_tokens": 794492737.0, "step": 20820 }, { "epoch": 2.648645210533011, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.217061042785645, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8854589462280273, "num_tokens": 794532589.0, "step": 20821 }, { "epoch": 2.6487724208116017, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.360713005065918, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.864380419254303, "num_tokens": 794566982.0, "step": 20822 }, { "epoch": 2.648899631090192, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.250828742980957, "learning_rate": 1e-06, "loss": 0.4459, "mean_token_accuracy": 0.8593541383743286, "num_tokens": 794604278.0, "step": 20823 }, { "epoch": 2.6490268413687827, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.302722930908203, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8843179941177368, "num_tokens": 794643785.0, "step": 20824 }, { "epoch": 2.649154051647373, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.283951759338379, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8849275708198547, "num_tokens": 794678302.0, "step": 20825 }, { "epoch": 2.649281261925964, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.298665046691895, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8646969795227051, "num_tokens": 794715732.0, "step": 20826 }, { "epoch": 2.649408472204554, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.279993057250977, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8794413805007935, "num_tokens": 794754359.0, "step": 20827 }, { "epoch": 2.649535682483145, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.306540489196777, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8706322908401489, "num_tokens": 794795770.0, "step": 20828 }, { "epoch": 2.649662892761735, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.3123197555542, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8681746125221252, "num_tokens": 794835654.0, "step": 20829 }, { "epoch": 2.649790103040326, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.298076629638672, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8771102428436279, "num_tokens": 794872527.0, "step": 20830 }, { "epoch": 2.649917313318916, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.250452041625977, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8783950805664062, "num_tokens": 794918577.0, "step": 20831 }, { "epoch": 2.6500445235975065, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.345834732055664, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8851439952850342, "num_tokens": 794953566.0, "step": 20832 }, { "epoch": 2.650171733876097, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.23495101928711, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8821471929550171, "num_tokens": 794990547.0, "step": 20833 }, { "epoch": 2.6502989441546876, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.266619682312012, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.891400933265686, "num_tokens": 795028419.0, "step": 20834 }, { "epoch": 2.650426154433278, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.276806831359863, "learning_rate": 1e-06, "loss": 0.3485, "mean_token_accuracy": 0.891066312789917, "num_tokens": 795064303.0, "step": 20835 }, { "epoch": 2.6505533647118686, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.319883346557617, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8542090058326721, "num_tokens": 795104280.0, "step": 20836 }, { "epoch": 2.650680574990459, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.276164054870605, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.866344153881073, "num_tokens": 795142547.0, "step": 20837 }, { "epoch": 2.6508077852690497, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.333222389221191, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.872599184513092, "num_tokens": 795186193.0, "step": 20838 }, { "epoch": 2.6509349955476402, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.262332916259766, "learning_rate": 1e-06, "loss": 0.3415, "mean_token_accuracy": 0.8918513059616089, "num_tokens": 795215453.0, "step": 20839 }, { "epoch": 2.6510622058262308, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.346095085144043, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8702831864356995, "num_tokens": 795252820.0, "step": 20840 }, { "epoch": 2.6511894161048213, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.293793678283691, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8717307448387146, "num_tokens": 795290000.0, "step": 20841 }, { "epoch": 2.651316626383412, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.344315528869629, "learning_rate": 1e-06, "loss": 0.3415, "mean_token_accuracy": 0.8924989104270935, "num_tokens": 795330389.0, "step": 20842 }, { "epoch": 2.6514438366620023, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25316333770752, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.872506856918335, "num_tokens": 795366693.0, "step": 20843 }, { "epoch": 2.651571046940593, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.311077117919922, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8763577342033386, "num_tokens": 795405748.0, "step": 20844 }, { "epoch": 2.6516982572191834, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.303547859191895, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8857390880584717, "num_tokens": 795446179.0, "step": 20845 }, { "epoch": 2.651825467497774, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.283785820007324, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8702654242515564, "num_tokens": 795488234.0, "step": 20846 }, { "epoch": 2.6519526777763645, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.252655982971191, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.872397780418396, "num_tokens": 795526508.0, "step": 20847 }, { "epoch": 2.652079888054955, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2952299118042, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.881172776222229, "num_tokens": 795566299.0, "step": 20848 }, { "epoch": 2.6522070983335455, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.23918342590332, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8708258867263794, "num_tokens": 795603892.0, "step": 20849 }, { "epoch": 2.6523343086121356, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.306870460510254, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.882675290107727, "num_tokens": 795639622.0, "step": 20850 }, { "epoch": 2.6524615188907266, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.271821022033691, "learning_rate": 1e-06, "loss": 0.3307, "mean_token_accuracy": 0.8958582878112793, "num_tokens": 795677299.0, "step": 20851 }, { "epoch": 2.6525887291693167, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.233716011047363, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.875328540802002, "num_tokens": 795714304.0, "step": 20852 }, { "epoch": 2.6527159394479076, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.220268249511719, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8707818388938904, "num_tokens": 795752330.0, "step": 20853 }, { "epoch": 2.6528431497264977, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.253387451171875, "learning_rate": 1e-06, "loss": 0.4571, "mean_token_accuracy": 0.8546357750892639, "num_tokens": 795789503.0, "step": 20854 }, { "epoch": 2.6529703600050887, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.244129180908203, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8733421564102173, "num_tokens": 795828449.0, "step": 20855 }, { "epoch": 2.6530975702836788, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.26714038848877, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8740711212158203, "num_tokens": 795868454.0, "step": 20856 }, { "epoch": 2.6532247805622693, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.225122451782227, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8625493049621582, "num_tokens": 795904629.0, "step": 20857 }, { "epoch": 2.65335199084086, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.280591011047363, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8612766265869141, "num_tokens": 795943197.0, "step": 20858 }, { "epoch": 2.6534792011194503, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.27621078491211, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8714002370834351, "num_tokens": 795984672.0, "step": 20859 }, { "epoch": 2.653606411398041, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.290457725524902, "learning_rate": 1e-06, "loss": 0.4536, "mean_token_accuracy": 0.8561902046203613, "num_tokens": 796022599.0, "step": 20860 }, { "epoch": 2.6537336216766314, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.257471084594727, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8727351427078247, "num_tokens": 796057798.0, "step": 20861 }, { "epoch": 2.653860831955222, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.284821510314941, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.873996376991272, "num_tokens": 796094928.0, "step": 20862 }, { "epoch": 2.6539880422338125, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.241148948669434, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8773304224014282, "num_tokens": 796132614.0, "step": 20863 }, { "epoch": 2.654115252512403, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.24079418182373, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8735059499740601, "num_tokens": 796171653.0, "step": 20864 }, { "epoch": 2.6542424627909935, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.20620346069336, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.8685558438301086, "num_tokens": 796215218.0, "step": 20865 }, { "epoch": 2.654369673069584, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.234333038330078, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8857119083404541, "num_tokens": 796249165.0, "step": 20866 }, { "epoch": 2.6544968833481746, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.262941360473633, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8716713190078735, "num_tokens": 796283878.0, "step": 20867 }, { "epoch": 2.654624093626765, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.229756355285645, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8623524904251099, "num_tokens": 796330760.0, "step": 20868 }, { "epoch": 2.6547513039053556, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.252381324768066, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8631606101989746, "num_tokens": 796371547.0, "step": 20869 }, { "epoch": 2.654878514183946, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.311427116394043, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8596186637878418, "num_tokens": 796409893.0, "step": 20870 }, { "epoch": 2.6550057244625367, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.259207725524902, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8593206405639648, "num_tokens": 796442493.0, "step": 20871 }, { "epoch": 2.655132934741127, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.30607795715332, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8769868016242981, "num_tokens": 796480853.0, "step": 20872 }, { "epoch": 2.6552601450197177, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.304339408874512, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8803108334541321, "num_tokens": 796520057.0, "step": 20873 }, { "epoch": 2.6553873552983083, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.24749755859375, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8788323998451233, "num_tokens": 796549779.0, "step": 20874 }, { "epoch": 2.6555145655768984, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.245431900024414, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8793759346008301, "num_tokens": 796590923.0, "step": 20875 }, { "epoch": 2.6556417758554893, "ewc_loss": 0.03857421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.287894248962402, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.869745135307312, "num_tokens": 796629455.0, "step": 20876 }, { "epoch": 2.6557689861340794, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25929069519043, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.870966911315918, "num_tokens": 796669950.0, "step": 20877 }, { "epoch": 2.6558961964126704, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.26797103881836, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8642407059669495, "num_tokens": 796712263.0, "step": 20878 }, { "epoch": 2.6560234066912605, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.312071800231934, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8743028044700623, "num_tokens": 796753025.0, "step": 20879 }, { "epoch": 2.6561506169698514, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.284452438354492, "learning_rate": 1e-06, "loss": 0.3537, "mean_token_accuracy": 0.8870115280151367, "num_tokens": 796788810.0, "step": 20880 }, { "epoch": 2.6562778272484415, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.264222145080566, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8766570091247559, "num_tokens": 796826275.0, "step": 20881 }, { "epoch": 2.656405037527032, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.265878677368164, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8687402606010437, "num_tokens": 796859587.0, "step": 20882 }, { "epoch": 2.6565322478056226, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.279017448425293, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8810275793075562, "num_tokens": 796899440.0, "step": 20883 }, { "epoch": 2.656659458084213, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.299127578735352, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8642858266830444, "num_tokens": 796938703.0, "step": 20884 }, { "epoch": 2.6567866683628036, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.252087593078613, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8608814477920532, "num_tokens": 796982287.0, "step": 20885 }, { "epoch": 2.656913878641394, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.317062377929688, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8762872219085693, "num_tokens": 797017640.0, "step": 20886 }, { "epoch": 2.6570410889199847, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.335153579711914, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8588269352912903, "num_tokens": 797054050.0, "step": 20887 }, { "epoch": 2.6571682991985752, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.278306007385254, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8784618377685547, "num_tokens": 797092429.0, "step": 20888 }, { "epoch": 2.6572955094771658, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.348166465759277, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8714337348937988, "num_tokens": 797125646.0, "step": 20889 }, { "epoch": 2.6574227197557563, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.258140563964844, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8774120807647705, "num_tokens": 797164560.0, "step": 20890 }, { "epoch": 2.657549930034347, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.4173002243042, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8728843927383423, "num_tokens": 797194168.0, "step": 20891 }, { "epoch": 2.6576771403129373, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.230639457702637, "learning_rate": 1e-06, "loss": 0.344, "mean_token_accuracy": 0.8928963541984558, "num_tokens": 797227045.0, "step": 20892 }, { "epoch": 2.657804350591528, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.254350662231445, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8883822560310364, "num_tokens": 797267607.0, "step": 20893 }, { "epoch": 2.6579315608701184, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25158405303955, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8635740280151367, "num_tokens": 797309320.0, "step": 20894 }, { "epoch": 2.658058771148709, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.33791732788086, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8715133666992188, "num_tokens": 797348258.0, "step": 20895 }, { "epoch": 2.6581859814272994, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.24927043914795, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8731200695037842, "num_tokens": 797391563.0, "step": 20896 }, { "epoch": 2.65831319170589, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.249796867370605, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8826539516448975, "num_tokens": 797428314.0, "step": 20897 }, { "epoch": 2.6584404019844805, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.359150886535645, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8539863228797913, "num_tokens": 797464480.0, "step": 20898 }, { "epoch": 2.658567612263071, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.281123161315918, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8759654760360718, "num_tokens": 797504135.0, "step": 20899 }, { "epoch": 2.658694822541661, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.275595664978027, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8771304488182068, "num_tokens": 797544222.0, "step": 20900 }, { "epoch": 2.658822032820252, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.248608589172363, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8797255754470825, "num_tokens": 797583695.0, "step": 20901 }, { "epoch": 2.658949243098842, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.287352561950684, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8606665730476379, "num_tokens": 797625660.0, "step": 20902 }, { "epoch": 2.659076453377433, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.254837036132812, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8734714984893799, "num_tokens": 797665807.0, "step": 20903 }, { "epoch": 2.6592036636560232, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.334685325622559, "learning_rate": 1e-06, "loss": 0.4508, "mean_token_accuracy": 0.8581757545471191, "num_tokens": 797707935.0, "step": 20904 }, { "epoch": 2.6593308739346138, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.321046829223633, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8676677942276001, "num_tokens": 797745502.0, "step": 20905 }, { "epoch": 2.6594580842132043, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.17524242401123, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8711338639259338, "num_tokens": 797779412.0, "step": 20906 }, { "epoch": 2.659585294491795, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.34066104888916, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8768179416656494, "num_tokens": 797816677.0, "step": 20907 }, { "epoch": 2.6597125047703853, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.258431434631348, "learning_rate": 1e-06, "loss": 0.4478, "mean_token_accuracy": 0.8564772009849548, "num_tokens": 797854260.0, "step": 20908 }, { "epoch": 2.659839715048976, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.265803337097168, "learning_rate": 1e-06, "loss": 0.4463, "mean_token_accuracy": 0.8613061904907227, "num_tokens": 797892349.0, "step": 20909 }, { "epoch": 2.6599669253275664, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.262540817260742, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8781749606132507, "num_tokens": 797928028.0, "step": 20910 }, { "epoch": 2.660094135606157, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.375548362731934, "learning_rate": 1e-06, "loss": 0.4447, "mean_token_accuracy": 0.8595068454742432, "num_tokens": 797963885.0, "step": 20911 }, { "epoch": 2.6602213458847475, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.325843811035156, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8615686893463135, "num_tokens": 798009842.0, "step": 20912 }, { "epoch": 2.660348556163338, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.308944702148438, "learning_rate": 1e-06, "loss": 0.3543, "mean_token_accuracy": 0.8863092660903931, "num_tokens": 798043207.0, "step": 20913 }, { "epoch": 2.6604757664419285, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.371171951293945, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8899927139282227, "num_tokens": 798080887.0, "step": 20914 }, { "epoch": 2.660602976720519, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.269972801208496, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8833109140396118, "num_tokens": 798126393.0, "step": 20915 }, { "epoch": 2.6607301869991096, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.312691688537598, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8893468976020813, "num_tokens": 798166099.0, "step": 20916 }, { "epoch": 2.6608573972777, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.213540077209473, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8678548336029053, "num_tokens": 798202285.0, "step": 20917 }, { "epoch": 2.6609846075562906, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.306900024414062, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8807715773582458, "num_tokens": 798243463.0, "step": 20918 }, { "epoch": 2.661111817834881, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.28684139251709, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8759275674819946, "num_tokens": 798275507.0, "step": 20919 }, { "epoch": 2.6612390281134717, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.26945686340332, "learning_rate": 1e-06, "loss": 0.3413, "mean_token_accuracy": 0.8919714093208313, "num_tokens": 798310536.0, "step": 20920 }, { "epoch": 2.661366238392062, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.331698417663574, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8694413900375366, "num_tokens": 798347874.0, "step": 20921 }, { "epoch": 2.6614934486706527, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.278088569641113, "learning_rate": 1e-06, "loss": 0.3261, "mean_token_accuracy": 0.896982729434967, "num_tokens": 798381865.0, "step": 20922 }, { "epoch": 2.661620658949243, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.326412200927734, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8725544214248657, "num_tokens": 798423378.0, "step": 20923 }, { "epoch": 2.661747869227834, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.284978866577148, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8802030086517334, "num_tokens": 798460278.0, "step": 20924 }, { "epoch": 2.661875079506424, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.31152057647705, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8758795261383057, "num_tokens": 798496622.0, "step": 20925 }, { "epoch": 2.662002289785015, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.334397315979004, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8698660731315613, "num_tokens": 798533061.0, "step": 20926 }, { "epoch": 2.662129500063605, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.259065628051758, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8665133118629456, "num_tokens": 798573689.0, "step": 20927 }, { "epoch": 2.662256710342196, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.346969604492188, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8836567401885986, "num_tokens": 798614058.0, "step": 20928 }, { "epoch": 2.662383920620786, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.219459533691406, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8638410568237305, "num_tokens": 798657769.0, "step": 20929 }, { "epoch": 2.6625111308993765, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.299822807312012, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.87744140625, "num_tokens": 798701084.0, "step": 20930 }, { "epoch": 2.662638341177967, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.30673885345459, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8799018859863281, "num_tokens": 798738029.0, "step": 20931 }, { "epoch": 2.6627655514565576, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.309080123901367, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8699563145637512, "num_tokens": 798773537.0, "step": 20932 }, { "epoch": 2.662892761735148, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.28953742980957, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8708111047744751, "num_tokens": 798810397.0, "step": 20933 }, { "epoch": 2.6630199720137386, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.352109909057617, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8705618381500244, "num_tokens": 798850796.0, "step": 20934 }, { "epoch": 2.663147182292329, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.33096981048584, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8733377456665039, "num_tokens": 798892379.0, "step": 20935 }, { "epoch": 2.6632743925709197, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.363557815551758, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.8620254397392273, "num_tokens": 798929444.0, "step": 20936 }, { "epoch": 2.66340160284951, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.350173950195312, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8812001943588257, "num_tokens": 798964086.0, "step": 20937 }, { "epoch": 2.6635288131281007, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.330635070800781, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8820183873176575, "num_tokens": 799007408.0, "step": 20938 }, { "epoch": 2.6636560234066913, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.36972427368164, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8788021206855774, "num_tokens": 799045930.0, "step": 20939 }, { "epoch": 2.663783233685282, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.295015335083008, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8839342594146729, "num_tokens": 799083568.0, "step": 20940 }, { "epoch": 2.6639104439638723, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.340271949768066, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.870705246925354, "num_tokens": 799123681.0, "step": 20941 }, { "epoch": 2.664037654242463, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.285211563110352, "learning_rate": 1e-06, "loss": 0.4164, "mean_token_accuracy": 0.8698172569274902, "num_tokens": 799164623.0, "step": 20942 }, { "epoch": 2.6641648645210534, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.364537239074707, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8735664486885071, "num_tokens": 799200408.0, "step": 20943 }, { "epoch": 2.664292074799644, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330562591552734, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8623204231262207, "num_tokens": 799245106.0, "step": 20944 }, { "epoch": 2.6644192850782344, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.297856330871582, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8812986612319946, "num_tokens": 799290429.0, "step": 20945 }, { "epoch": 2.664546495356825, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.354494094848633, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8725324869155884, "num_tokens": 799322951.0, "step": 20946 }, { "epoch": 2.6646737056354155, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.26968765258789, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.870993971824646, "num_tokens": 799365498.0, "step": 20947 }, { "epoch": 2.6648009159140056, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.355710983276367, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8684613704681396, "num_tokens": 799405894.0, "step": 20948 }, { "epoch": 2.6649281261925966, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.315895080566406, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8733870387077332, "num_tokens": 799439750.0, "step": 20949 }, { "epoch": 2.6650553364711866, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.332284927368164, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8658050894737244, "num_tokens": 799475170.0, "step": 20950 }, { "epoch": 2.6651825467497776, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.33609390258789, "learning_rate": 1e-06, "loss": 0.446, "mean_token_accuracy": 0.8611942529678345, "num_tokens": 799514231.0, "step": 20951 }, { "epoch": 2.6653097570283677, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.279701232910156, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8743171095848083, "num_tokens": 799549512.0, "step": 20952 }, { "epoch": 2.6654369673069587, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.317985534667969, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8689565658569336, "num_tokens": 799592144.0, "step": 20953 }, { "epoch": 2.6655641775855488, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.323041915893555, "learning_rate": 1e-06, "loss": 0.4426, "mean_token_accuracy": 0.8595330715179443, "num_tokens": 799634779.0, "step": 20954 }, { "epoch": 2.6656913878641393, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.254029273986816, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8857877254486084, "num_tokens": 799677548.0, "step": 20955 }, { "epoch": 2.66581859814273, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.3425874710083, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8566049337387085, "num_tokens": 799715679.0, "step": 20956 }, { "epoch": 2.6659458084213203, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.338043212890625, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.866703987121582, "num_tokens": 799755701.0, "step": 20957 }, { "epoch": 2.666073018699911, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.363944053649902, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.869443416595459, "num_tokens": 799795151.0, "step": 20958 }, { "epoch": 2.6662002289785014, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.288134574890137, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8805668950080872, "num_tokens": 799832249.0, "step": 20959 }, { "epoch": 2.666327439257092, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.339350700378418, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8536670207977295, "num_tokens": 799867762.0, "step": 20960 }, { "epoch": 2.6664546495356825, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.358904838562012, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8708827495574951, "num_tokens": 799909012.0, "step": 20961 }, { "epoch": 2.666581859814273, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.288246154785156, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8706855773925781, "num_tokens": 799950316.0, "step": 20962 }, { "epoch": 2.6667090700928635, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.293252944946289, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.8881505727767944, "num_tokens": 799990037.0, "step": 20963 }, { "epoch": 2.666836280371454, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.383285522460938, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8828788995742798, "num_tokens": 800023935.0, "step": 20964 }, { "epoch": 2.6669634906500446, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.25234603881836, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8795344829559326, "num_tokens": 800066357.0, "step": 20965 }, { "epoch": 2.667090700928635, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.330504417419434, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8604415655136108, "num_tokens": 800100667.0, "step": 20966 }, { "epoch": 2.6672179112072256, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.347271919250488, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8774969577789307, "num_tokens": 800135101.0, "step": 20967 }, { "epoch": 2.667345121485816, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.30102252960205, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8877370357513428, "num_tokens": 800176308.0, "step": 20968 }, { "epoch": 2.6674723317644067, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.361479759216309, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.884986400604248, "num_tokens": 800205585.0, "step": 20969 }, { "epoch": 2.667599542042997, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.344522476196289, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8728379011154175, "num_tokens": 800241400.0, "step": 20970 }, { "epoch": 2.6677267523215877, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.307883262634277, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8836694955825806, "num_tokens": 800279060.0, "step": 20971 }, { "epoch": 2.6678539626001783, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.363566398620605, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8830692768096924, "num_tokens": 800318606.0, "step": 20972 }, { "epoch": 2.6679811728787683, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.31599235534668, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8725262880325317, "num_tokens": 800356795.0, "step": 20973 }, { "epoch": 2.6681083831573593, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330328941345215, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.881316065788269, "num_tokens": 800393786.0, "step": 20974 }, { "epoch": 2.6682355934359494, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.326087951660156, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8724100589752197, "num_tokens": 800440153.0, "step": 20975 }, { "epoch": 2.6683628037145404, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.339988708496094, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8559863567352295, "num_tokens": 800476343.0, "step": 20976 }, { "epoch": 2.6684900139931305, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.28011703491211, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8860656023025513, "num_tokens": 800514227.0, "step": 20977 }, { "epoch": 2.668617224271721, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.3595552444458, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8721685409545898, "num_tokens": 800552357.0, "step": 20978 }, { "epoch": 2.6687444345503115, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.309311866760254, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8813368082046509, "num_tokens": 800587624.0, "step": 20979 }, { "epoch": 2.668871644828902, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.26426887512207, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8680139780044556, "num_tokens": 800631183.0, "step": 20980 }, { "epoch": 2.6689988551074926, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.33934497833252, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8781760931015015, "num_tokens": 800670334.0, "step": 20981 }, { "epoch": 2.669126065386083, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.348897933959961, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8706537485122681, "num_tokens": 800703500.0, "step": 20982 }, { "epoch": 2.6692532756646736, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.30901050567627, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8671626448631287, "num_tokens": 800742904.0, "step": 20983 }, { "epoch": 2.669380485943264, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.314596176147461, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8712825775146484, "num_tokens": 800777556.0, "step": 20984 }, { "epoch": 2.6695076962218547, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.356054306030273, "learning_rate": 1e-06, "loss": 0.4187, "mean_token_accuracy": 0.8653795719146729, "num_tokens": 800817532.0, "step": 20985 }, { "epoch": 2.669634906500445, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.252359390258789, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8835434913635254, "num_tokens": 800856567.0, "step": 20986 }, { "epoch": 2.6697621167790357, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.391139030456543, "learning_rate": 1e-06, "loss": 0.4638, "mean_token_accuracy": 0.8520053029060364, "num_tokens": 800901234.0, "step": 20987 }, { "epoch": 2.6698893270576263, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.253268241882324, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8712188005447388, "num_tokens": 800942647.0, "step": 20988 }, { "epoch": 2.670016537336217, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.343952178955078, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.880893886089325, "num_tokens": 800980194.0, "step": 20989 }, { "epoch": 2.6701437476148073, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.275368690490723, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8789185285568237, "num_tokens": 801016978.0, "step": 20990 }, { "epoch": 2.670270957893398, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.377607345581055, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8601755499839783, "num_tokens": 801051921.0, "step": 20991 }, { "epoch": 2.6703981681719884, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.249924659729004, "learning_rate": 1e-06, "loss": 0.4666, "mean_token_accuracy": 0.8484929203987122, "num_tokens": 801087189.0, "step": 20992 }, { "epoch": 2.670525378450579, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.362037658691406, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8780664801597595, "num_tokens": 801121574.0, "step": 20993 }, { "epoch": 2.6706525887291694, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.249218940734863, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8791041374206543, "num_tokens": 801160497.0, "step": 20994 }, { "epoch": 2.67077979900776, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.312521934509277, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8743294477462769, "num_tokens": 801198556.0, "step": 20995 }, { "epoch": 2.6709070092863505, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.294426918029785, "learning_rate": 1e-06, "loss": 0.3204, "mean_token_accuracy": 0.8987497687339783, "num_tokens": 801230616.0, "step": 20996 }, { "epoch": 2.671034219564941, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.273681640625, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8721268177032471, "num_tokens": 801266582.0, "step": 20997 }, { "epoch": 2.671161429843531, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.283177375793457, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8850446939468384, "num_tokens": 801305127.0, "step": 20998 }, { "epoch": 2.671288640122122, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.222407341003418, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8788535594940186, "num_tokens": 801348284.0, "step": 20999 }, { "epoch": 2.671415850400712, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.300626754760742, "learning_rate": 1e-06, "loss": 0.3375, "mean_token_accuracy": 0.8933061957359314, "num_tokens": 801385298.0, "step": 21000 }, { "epoch": 2.671543060679303, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.322497367858887, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8660218715667725, "num_tokens": 801417020.0, "step": 21001 }, { "epoch": 2.6716702709578932, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.296370506286621, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8664418458938599, "num_tokens": 801455328.0, "step": 21002 }, { "epoch": 2.6717974812364838, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.270548820495605, "learning_rate": 1e-06, "loss": 0.3673, "mean_token_accuracy": 0.8839842081069946, "num_tokens": 801494342.0, "step": 21003 }, { "epoch": 2.6719246915150743, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.245278358459473, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8760244846343994, "num_tokens": 801528182.0, "step": 21004 }, { "epoch": 2.672051901793665, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.299966812133789, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8742073774337769, "num_tokens": 801569424.0, "step": 21005 }, { "epoch": 2.6721791120722553, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.29745101928711, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8725268840789795, "num_tokens": 801608809.0, "step": 21006 }, { "epoch": 2.672306322350846, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.27754020690918, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8653242588043213, "num_tokens": 801651910.0, "step": 21007 }, { "epoch": 2.6724335326294364, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.310419082641602, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8726016283035278, "num_tokens": 801692961.0, "step": 21008 }, { "epoch": 2.672560742908027, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.292781829833984, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8615280389785767, "num_tokens": 801731278.0, "step": 21009 }, { "epoch": 2.6726879531866174, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.293448448181152, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.870421826839447, "num_tokens": 801769562.0, "step": 21010 }, { "epoch": 2.672815163465208, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.303619384765625, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8691955804824829, "num_tokens": 801811046.0, "step": 21011 }, { "epoch": 2.6729423737437985, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.277053833007812, "learning_rate": 1e-06, "loss": 0.466, "mean_token_accuracy": 0.8515801429748535, "num_tokens": 801847833.0, "step": 21012 }, { "epoch": 2.673069584022389, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.2940673828125, "learning_rate": 1e-06, "loss": 0.4235, "mean_token_accuracy": 0.868333637714386, "num_tokens": 801887819.0, "step": 21013 }, { "epoch": 2.6731967943009796, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.362399101257324, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8848361372947693, "num_tokens": 801922535.0, "step": 21014 }, { "epoch": 2.67332400457957, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.305800437927246, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8782742023468018, "num_tokens": 801958025.0, "step": 21015 }, { "epoch": 2.6734512148581606, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.34663200378418, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8782879114151001, "num_tokens": 801991291.0, "step": 21016 }, { "epoch": 2.673578425136751, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.294134140014648, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.875461220741272, "num_tokens": 802024952.0, "step": 21017 }, { "epoch": 2.6737056354153417, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.292943000793457, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8729351758956909, "num_tokens": 802065741.0, "step": 21018 }, { "epoch": 2.673832845693932, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.26547622680664, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8728718161582947, "num_tokens": 802107153.0, "step": 21019 }, { "epoch": 2.6739600559725227, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.291919708251953, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8792715072631836, "num_tokens": 802148177.0, "step": 21020 }, { "epoch": 2.674087266251113, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.305845260620117, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8755413293838501, "num_tokens": 802186144.0, "step": 21021 }, { "epoch": 2.674214476529704, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.353628158569336, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.8876999616622925, "num_tokens": 802224054.0, "step": 21022 }, { "epoch": 2.674341686808294, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.320324897766113, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.85743647813797, "num_tokens": 802264483.0, "step": 21023 }, { "epoch": 2.674468897086885, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.253022193908691, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8813186883926392, "num_tokens": 802302933.0, "step": 21024 }, { "epoch": 2.674596107365475, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.385193824768066, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8669623136520386, "num_tokens": 802347354.0, "step": 21025 }, { "epoch": 2.674723317644066, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.250568389892578, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8825994729995728, "num_tokens": 802392159.0, "step": 21026 }, { "epoch": 2.674850527922656, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.264447212219238, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8775956630706787, "num_tokens": 802432629.0, "step": 21027 }, { "epoch": 2.6749777382012465, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.316110610961914, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8828393816947937, "num_tokens": 802471776.0, "step": 21028 }, { "epoch": 2.675104948479837, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.326409339904785, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8675802946090698, "num_tokens": 802507314.0, "step": 21029 }, { "epoch": 2.6752321587584276, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330587387084961, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8815773129463196, "num_tokens": 802550587.0, "step": 21030 }, { "epoch": 2.675359369037018, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.259873390197754, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8743969202041626, "num_tokens": 802591544.0, "step": 21031 }, { "epoch": 2.6754865793156086, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.31650161743164, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8810479640960693, "num_tokens": 802627336.0, "step": 21032 }, { "epoch": 2.675613789594199, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.311015129089355, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8728220462799072, "num_tokens": 802668358.0, "step": 21033 }, { "epoch": 2.6757409998727897, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.319145202636719, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8809127807617188, "num_tokens": 802708023.0, "step": 21034 }, { "epoch": 2.67586821015138, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.311635971069336, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.8944498896598816, "num_tokens": 802745856.0, "step": 21035 }, { "epoch": 2.6759954204299707, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.287995338439941, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8854195475578308, "num_tokens": 802779032.0, "step": 21036 }, { "epoch": 2.6761226307085613, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.345392227172852, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8741558194160461, "num_tokens": 802815736.0, "step": 21037 }, { "epoch": 2.676249840987152, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.342098236083984, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.8706992864608765, "num_tokens": 802854133.0, "step": 21038 }, { "epoch": 2.6763770512657423, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.33323860168457, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8808311223983765, "num_tokens": 802892969.0, "step": 21039 }, { "epoch": 2.676504261544333, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.3325834274292, "learning_rate": 1e-06, "loss": 0.3559, "mean_token_accuracy": 0.8890057802200317, "num_tokens": 802936998.0, "step": 21040 }, { "epoch": 2.6766314718229234, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.311829566955566, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8830548524856567, "num_tokens": 802979040.0, "step": 21041 }, { "epoch": 2.676758682101514, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.314949035644531, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8852860927581787, "num_tokens": 803016516.0, "step": 21042 }, { "epoch": 2.6768858923801044, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.337386131286621, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.872492790222168, "num_tokens": 803055566.0, "step": 21043 }, { "epoch": 2.677013102658695, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.281061172485352, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8763971328735352, "num_tokens": 803093683.0, "step": 21044 }, { "epoch": 2.6771403129372855, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.349682807922363, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8833520412445068, "num_tokens": 803133691.0, "step": 21045 }, { "epoch": 2.6772675232158756, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.301714897155762, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8833056092262268, "num_tokens": 803169870.0, "step": 21046 }, { "epoch": 2.6773947334944665, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.369377136230469, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8673795461654663, "num_tokens": 803209038.0, "step": 21047 }, { "epoch": 2.6775219437730566, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.363078117370605, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8606056571006775, "num_tokens": 803239725.0, "step": 21048 }, { "epoch": 2.6776491540516476, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.382704734802246, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8728988170623779, "num_tokens": 803273580.0, "step": 21049 }, { "epoch": 2.6777763643302377, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.286038398742676, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8799375891685486, "num_tokens": 803313583.0, "step": 21050 }, { "epoch": 2.6779035746088287, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.338937759399414, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.883745551109314, "num_tokens": 803347517.0, "step": 21051 }, { "epoch": 2.6780307848874187, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.344162940979004, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8611930012702942, "num_tokens": 803388183.0, "step": 21052 }, { "epoch": 2.6781579951660093, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.33659839630127, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.874690592288971, "num_tokens": 803428377.0, "step": 21053 }, { "epoch": 2.6782852054446, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.353516578674316, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.883064329624176, "num_tokens": 803460930.0, "step": 21054 }, { "epoch": 2.6784124157231903, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.314537048339844, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8828664422035217, "num_tokens": 803497971.0, "step": 21055 }, { "epoch": 2.678539626001781, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.352194786071777, "learning_rate": 1e-06, "loss": 0.3521, "mean_token_accuracy": 0.8896186351776123, "num_tokens": 803535514.0, "step": 21056 }, { "epoch": 2.6786668362803714, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.266054153442383, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8711637258529663, "num_tokens": 803580767.0, "step": 21057 }, { "epoch": 2.678794046558962, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.348240852355957, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.877810537815094, "num_tokens": 803619168.0, "step": 21058 }, { "epoch": 2.6789212568375524, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.336326599121094, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8621349334716797, "num_tokens": 803654785.0, "step": 21059 }, { "epoch": 2.679048467116143, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.353302001953125, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8827050924301147, "num_tokens": 803692505.0, "step": 21060 }, { "epoch": 2.6791756773947335, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.377756118774414, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8823316097259521, "num_tokens": 803732592.0, "step": 21061 }, { "epoch": 2.679302887673324, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.33540153503418, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8631319403648376, "num_tokens": 803767601.0, "step": 21062 }, { "epoch": 2.6794300979519146, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.28328800201416, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.868054211139679, "num_tokens": 803804329.0, "step": 21063 }, { "epoch": 2.679557308230505, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.339927673339844, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8706207275390625, "num_tokens": 803846182.0, "step": 21064 }, { "epoch": 2.6796845185090956, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.34190845489502, "learning_rate": 1e-06, "loss": 0.4243, "mean_token_accuracy": 0.8703615069389343, "num_tokens": 803883561.0, "step": 21065 }, { "epoch": 2.679811728787686, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.352615356445312, "learning_rate": 1e-06, "loss": 0.3414, "mean_token_accuracy": 0.8932203054428101, "num_tokens": 803916860.0, "step": 21066 }, { "epoch": 2.6799389390662767, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330155372619629, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.874880850315094, "num_tokens": 803951949.0, "step": 21067 }, { "epoch": 2.680066149344867, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.336298942565918, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8710360527038574, "num_tokens": 803992094.0, "step": 21068 }, { "epoch": 2.6801933596234577, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.305171966552734, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8744332790374756, "num_tokens": 804034058.0, "step": 21069 }, { "epoch": 2.6803205699020483, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.33505630493164, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.867155134677887, "num_tokens": 804072722.0, "step": 21070 }, { "epoch": 2.6804477801806383, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.298274040222168, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8798007965087891, "num_tokens": 804114253.0, "step": 21071 }, { "epoch": 2.6805749904592293, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.319340705871582, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8771889805793762, "num_tokens": 804154877.0, "step": 21072 }, { "epoch": 2.6807022007378194, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.371355056762695, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8698563575744629, "num_tokens": 804191060.0, "step": 21073 }, { "epoch": 2.6808294110164104, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.324023246765137, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.866452693939209, "num_tokens": 804231574.0, "step": 21074 }, { "epoch": 2.6809566212950005, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.36161994934082, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8538293838500977, "num_tokens": 804264027.0, "step": 21075 }, { "epoch": 2.681083831573591, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.308884620666504, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8757452368736267, "num_tokens": 804302906.0, "step": 21076 }, { "epoch": 2.6812110418521815, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.39089298248291, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8701062798500061, "num_tokens": 804339107.0, "step": 21077 }, { "epoch": 2.681338252130772, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.332067489624023, "learning_rate": 1e-06, "loss": 0.3857, "mean_token_accuracy": 0.8794682025909424, "num_tokens": 804374292.0, "step": 21078 }, { "epoch": 2.6814654624093626, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.298566818237305, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8760644197463989, "num_tokens": 804420287.0, "step": 21079 }, { "epoch": 2.681592672687953, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.365878105163574, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8796353340148926, "num_tokens": 804453977.0, "step": 21080 }, { "epoch": 2.6817198829665436, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.36090087890625, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8594328165054321, "num_tokens": 804494958.0, "step": 21081 }, { "epoch": 2.681847093245134, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.307672500610352, "learning_rate": 1e-06, "loss": 0.4506, "mean_token_accuracy": 0.8549972772598267, "num_tokens": 804532209.0, "step": 21082 }, { "epoch": 2.6819743035237247, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.31779956817627, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8775549530982971, "num_tokens": 804572937.0, "step": 21083 }, { "epoch": 2.682101513802315, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.297693252563477, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8754139542579651, "num_tokens": 804610263.0, "step": 21084 }, { "epoch": 2.6822287240809057, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.392940521240234, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.879146158695221, "num_tokens": 804645516.0, "step": 21085 }, { "epoch": 2.6823559343594963, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.295221328735352, "learning_rate": 1e-06, "loss": 0.3745, "mean_token_accuracy": 0.8856674432754517, "num_tokens": 804684540.0, "step": 21086 }, { "epoch": 2.682483144638087, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.333648681640625, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8759588003158569, "num_tokens": 804723030.0, "step": 21087 }, { "epoch": 2.6826103549166773, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.301268577575684, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8829430937767029, "num_tokens": 804759202.0, "step": 21088 }, { "epoch": 2.682737565195268, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.383771896362305, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8715838193893433, "num_tokens": 804803199.0, "step": 21089 }, { "epoch": 2.6828647754738584, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.285167694091797, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8727436065673828, "num_tokens": 804835675.0, "step": 21090 }, { "epoch": 2.682991985752449, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.341714859008789, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8625541925430298, "num_tokens": 804876583.0, "step": 21091 }, { "epoch": 2.6831191960310394, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.353413581848145, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8651532530784607, "num_tokens": 804906110.0, "step": 21092 }, { "epoch": 2.68324640630963, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.386164665222168, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.8659083843231201, "num_tokens": 804942196.0, "step": 21093 }, { "epoch": 2.6833736165882205, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.294463157653809, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8745925426483154, "num_tokens": 804978760.0, "step": 21094 }, { "epoch": 2.683500826866811, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.414748191833496, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.867131233215332, "num_tokens": 805014212.0, "step": 21095 }, { "epoch": 2.683628037145401, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.310839653015137, "learning_rate": 1e-06, "loss": 0.3284, "mean_token_accuracy": 0.8964759707450867, "num_tokens": 805050528.0, "step": 21096 }, { "epoch": 2.683755247423992, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.382962226867676, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.8750525116920471, "num_tokens": 805087355.0, "step": 21097 }, { "epoch": 2.683882457702582, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.372800827026367, "learning_rate": 1e-06, "loss": 0.4356, "mean_token_accuracy": 0.8598212003707886, "num_tokens": 805117837.0, "step": 21098 }, { "epoch": 2.684009667981173, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.350287437438965, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8739374876022339, "num_tokens": 805157734.0, "step": 21099 }, { "epoch": 2.684136878259763, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.292360305786133, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.8811473250389099, "num_tokens": 805190657.0, "step": 21100 }, { "epoch": 2.6842640885383537, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.389141082763672, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8683050870895386, "num_tokens": 805229654.0, "step": 21101 }, { "epoch": 2.6843912988169443, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.372873306274414, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8749446272850037, "num_tokens": 805269199.0, "step": 21102 }, { "epoch": 2.684518509095535, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.329890251159668, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8692178130149841, "num_tokens": 805308918.0, "step": 21103 }, { "epoch": 2.6846457193741253, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.384626388549805, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8711912631988525, "num_tokens": 805345014.0, "step": 21104 }, { "epoch": 2.684772929652716, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.358792304992676, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8864186406135559, "num_tokens": 805383674.0, "step": 21105 }, { "epoch": 2.6849001399313064, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.340373039245605, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.878494143486023, "num_tokens": 805419765.0, "step": 21106 }, { "epoch": 2.685027350209897, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.32233715057373, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8749595880508423, "num_tokens": 805461346.0, "step": 21107 }, { "epoch": 2.6851545604884874, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.366726875305176, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8816514015197754, "num_tokens": 805494519.0, "step": 21108 }, { "epoch": 2.685281770767078, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.344693183898926, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8641082048416138, "num_tokens": 805530542.0, "step": 21109 }, { "epoch": 2.6854089810456685, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.326715469360352, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8727887868881226, "num_tokens": 805564531.0, "step": 21110 }, { "epoch": 2.685536191324259, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330897331237793, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8707285523414612, "num_tokens": 805606011.0, "step": 21111 }, { "epoch": 2.6856634016028496, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.410313606262207, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8680238127708435, "num_tokens": 805643078.0, "step": 21112 }, { "epoch": 2.68579061188144, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.284489631652832, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8775709867477417, "num_tokens": 805678554.0, "step": 21113 }, { "epoch": 2.6859178221600306, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.35152816772461, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8837459683418274, "num_tokens": 805713736.0, "step": 21114 }, { "epoch": 2.686045032438621, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.361471176147461, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8681678771972656, "num_tokens": 805751798.0, "step": 21115 }, { "epoch": 2.6861722427172117, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.307284355163574, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8812111020088196, "num_tokens": 805789205.0, "step": 21116 }, { "epoch": 2.686299452995802, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.360055923461914, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8782891035079956, "num_tokens": 805824910.0, "step": 21117 }, { "epoch": 2.6864266632743927, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.325763702392578, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.872796893119812, "num_tokens": 805861335.0, "step": 21118 }, { "epoch": 2.686553873552983, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.363907814025879, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8708128929138184, "num_tokens": 805901488.0, "step": 21119 }, { "epoch": 2.686681083831574, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.370390892028809, "learning_rate": 1e-06, "loss": 0.5071, "mean_token_accuracy": 0.8370945453643799, "num_tokens": 805941336.0, "step": 21120 }, { "epoch": 2.686808294110164, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.362052917480469, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8818427324295044, "num_tokens": 805974222.0, "step": 21121 }, { "epoch": 2.686935504388755, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.345407485961914, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8706717491149902, "num_tokens": 806014942.0, "step": 21122 }, { "epoch": 2.687062714667345, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.314311981201172, "learning_rate": 1e-06, "loss": 0.36, "mean_token_accuracy": 0.8840516805648804, "num_tokens": 806048894.0, "step": 21123 }, { "epoch": 2.687189924945936, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.399751663208008, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.872409999370575, "num_tokens": 806089714.0, "step": 21124 }, { "epoch": 2.687317135224526, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.290519714355469, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.862580418586731, "num_tokens": 806128974.0, "step": 21125 }, { "epoch": 2.6874443455031165, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.371230125427246, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8733307123184204, "num_tokens": 806169857.0, "step": 21126 }, { "epoch": 2.687571555781707, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.317849159240723, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8753159046173096, "num_tokens": 806202826.0, "step": 21127 }, { "epoch": 2.6876987660602976, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.335047721862793, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8721983432769775, "num_tokens": 806242850.0, "step": 21128 }, { "epoch": 2.687825976338888, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.23766040802002, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8757456541061401, "num_tokens": 806279024.0, "step": 21129 }, { "epoch": 2.6879531866174786, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.39497184753418, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.860281765460968, "num_tokens": 806313669.0, "step": 21130 }, { "epoch": 2.688080396896069, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.312093734741211, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8776892423629761, "num_tokens": 806353128.0, "step": 21131 }, { "epoch": 2.6882076071746597, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.403759956359863, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8755548000335693, "num_tokens": 806385044.0, "step": 21132 }, { "epoch": 2.68833481745325, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.355326652526855, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.865756630897522, "num_tokens": 806422157.0, "step": 21133 }, { "epoch": 2.6884620277318407, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.335681915283203, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8734327554702759, "num_tokens": 806463180.0, "step": 21134 }, { "epoch": 2.6885892380104313, "ewc_loss": 0.038818359375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 14.404114723205566, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8700898289680481, "num_tokens": 806502031.0, "step": 21135 }, { "epoch": 2.688716448289022, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.326053619384766, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8756923675537109, "num_tokens": 806539585.0, "step": 21136 }, { "epoch": 2.6888436585676123, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.405972480773926, "learning_rate": 1e-06, "loss": 0.3404, "mean_token_accuracy": 0.8941980004310608, "num_tokens": 806573147.0, "step": 21137 }, { "epoch": 2.688970868846203, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.262872695922852, "learning_rate": 1e-06, "loss": 0.3617, "mean_token_accuracy": 0.8865670561790466, "num_tokens": 806609785.0, "step": 21138 }, { "epoch": 2.6890980791247934, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.34979248046875, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.881496250629425, "num_tokens": 806647709.0, "step": 21139 }, { "epoch": 2.689225289403384, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.32200813293457, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8782714605331421, "num_tokens": 806684448.0, "step": 21140 }, { "epoch": 2.6893524996819744, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.315673828125, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8780006170272827, "num_tokens": 806725326.0, "step": 21141 }, { "epoch": 2.689479709960565, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.332700729370117, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8736052513122559, "num_tokens": 806759797.0, "step": 21142 }, { "epoch": 2.6896069202391555, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.304863929748535, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8807213306427002, "num_tokens": 806799986.0, "step": 21143 }, { "epoch": 2.6897341305177456, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.350666046142578, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8687952756881714, "num_tokens": 806843055.0, "step": 21144 }, { "epoch": 2.6898613407963365, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.335607528686523, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8884143233299255, "num_tokens": 806875897.0, "step": 21145 }, { "epoch": 2.6899885510749266, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.331133842468262, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8822824954986572, "num_tokens": 806915583.0, "step": 21146 }, { "epoch": 2.6901157613535176, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.366788864135742, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8770750761032104, "num_tokens": 806953851.0, "step": 21147 }, { "epoch": 2.6902429716321077, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.3605318069458, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.870861291885376, "num_tokens": 806988519.0, "step": 21148 }, { "epoch": 2.6903701819106987, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.359565734863281, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8775518536567688, "num_tokens": 807029612.0, "step": 21149 }, { "epoch": 2.6904973921892887, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330163955688477, "learning_rate": 1e-06, "loss": 0.4442, "mean_token_accuracy": 0.8593009114265442, "num_tokens": 807064009.0, "step": 21150 }, { "epoch": 2.6906246024678793, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.349137306213379, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8722858428955078, "num_tokens": 807104694.0, "step": 21151 }, { "epoch": 2.69075181274647, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.345852851867676, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8528352975845337, "num_tokens": 807142500.0, "step": 21152 }, { "epoch": 2.6908790230250603, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.372629165649414, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8709223866462708, "num_tokens": 807172123.0, "step": 21153 }, { "epoch": 2.691006233303651, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.345023155212402, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8758373260498047, "num_tokens": 807206422.0, "step": 21154 }, { "epoch": 2.6911334435822414, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.351813316345215, "learning_rate": 1e-06, "loss": 0.4323, "mean_token_accuracy": 0.8602533936500549, "num_tokens": 807240867.0, "step": 21155 }, { "epoch": 2.691260653860832, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.33316421508789, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8832862377166748, "num_tokens": 807275838.0, "step": 21156 }, { "epoch": 2.6913878641394224, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.326092720031738, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8827263116836548, "num_tokens": 807315164.0, "step": 21157 }, { "epoch": 2.691515074418013, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.365970611572266, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8649709820747375, "num_tokens": 807358016.0, "step": 21158 }, { "epoch": 2.6916422846966035, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.297942161560059, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.873131513595581, "num_tokens": 807402751.0, "step": 21159 }, { "epoch": 2.691769494975194, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.332663536071777, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8782893419265747, "num_tokens": 807442761.0, "step": 21160 }, { "epoch": 2.6918967052537845, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.351317405700684, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8812887072563171, "num_tokens": 807478550.0, "step": 21161 }, { "epoch": 2.692023915532375, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.34310245513916, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.883082389831543, "num_tokens": 807522756.0, "step": 21162 }, { "epoch": 2.6921511258109656, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.34559440612793, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8789823055267334, "num_tokens": 807561701.0, "step": 21163 }, { "epoch": 2.692278336089556, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.372942924499512, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8731446862220764, "num_tokens": 807604058.0, "step": 21164 }, { "epoch": 2.6924055463681467, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.390313148498535, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8797996044158936, "num_tokens": 807637344.0, "step": 21165 }, { "epoch": 2.692532756646737, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.359158515930176, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.866997480392456, "num_tokens": 807676867.0, "step": 21166 }, { "epoch": 2.6926599669253277, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.384605407714844, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.877205491065979, "num_tokens": 807707183.0, "step": 21167 }, { "epoch": 2.6927871772039182, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.377426147460938, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8661679029464722, "num_tokens": 807737617.0, "step": 21168 }, { "epoch": 2.6929143874825083, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.320804595947266, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8663676381111145, "num_tokens": 807774141.0, "step": 21169 }, { "epoch": 2.6930415977610993, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.362289428710938, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8795596361160278, "num_tokens": 807813375.0, "step": 21170 }, { "epoch": 2.6931688080396894, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330021858215332, "learning_rate": 1e-06, "loss": 0.4383, "mean_token_accuracy": 0.86216801404953, "num_tokens": 807852317.0, "step": 21171 }, { "epoch": 2.6932960183182804, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.36596965789795, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8759498596191406, "num_tokens": 807895266.0, "step": 21172 }, { "epoch": 2.6934232285968704, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.356189727783203, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8758280277252197, "num_tokens": 807938600.0, "step": 21173 }, { "epoch": 2.693550438875461, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.356762886047363, "learning_rate": 1e-06, "loss": 0.3375, "mean_token_accuracy": 0.8949711322784424, "num_tokens": 807974329.0, "step": 21174 }, { "epoch": 2.6936776491540515, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.365196228027344, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8862822651863098, "num_tokens": 808015666.0, "step": 21175 }, { "epoch": 2.693804859432642, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.329508781433105, "learning_rate": 1e-06, "loss": 0.3692, "mean_token_accuracy": 0.8850837349891663, "num_tokens": 808050176.0, "step": 21176 }, { "epoch": 2.6939320697112326, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.345369338989258, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8672367334365845, "num_tokens": 808091910.0, "step": 21177 }, { "epoch": 2.694059279989823, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.316720008850098, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8696550130844116, "num_tokens": 808128646.0, "step": 21178 }, { "epoch": 2.6941864902684136, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.389521598815918, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8783779144287109, "num_tokens": 808166544.0, "step": 21179 }, { "epoch": 2.694313700547004, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.322684288024902, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8634710907936096, "num_tokens": 808200051.0, "step": 21180 }, { "epoch": 2.6944409108255947, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.403292655944824, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8757377862930298, "num_tokens": 808239251.0, "step": 21181 }, { "epoch": 2.694568121104185, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.319915771484375, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8688187003135681, "num_tokens": 808280639.0, "step": 21182 }, { "epoch": 2.6946953313827757, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.382951736450195, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8806090354919434, "num_tokens": 808316141.0, "step": 21183 }, { "epoch": 2.6948225416613663, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.327666282653809, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8727671504020691, "num_tokens": 808359172.0, "step": 21184 }, { "epoch": 2.694949751939957, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.394591331481934, "learning_rate": 1e-06, "loss": 0.4923, "mean_token_accuracy": 0.8464616537094116, "num_tokens": 808397823.0, "step": 21185 }, { "epoch": 2.6950769622185473, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.361438751220703, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8837478160858154, "num_tokens": 808435355.0, "step": 21186 }, { "epoch": 2.695204172497138, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.376566886901855, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8805127143859863, "num_tokens": 808470792.0, "step": 21187 }, { "epoch": 2.6953313827757284, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.336649894714355, "learning_rate": 1e-06, "loss": 0.3462, "mean_token_accuracy": 0.8911576867103577, "num_tokens": 808504591.0, "step": 21188 }, { "epoch": 2.695458593054319, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.448083877563477, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8853488564491272, "num_tokens": 808537167.0, "step": 21189 }, { "epoch": 2.6955858033329094, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.373712539672852, "learning_rate": 1e-06, "loss": 0.3537, "mean_token_accuracy": 0.8888363838195801, "num_tokens": 808572055.0, "step": 21190 }, { "epoch": 2.6957130136115, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.31101131439209, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8862167000770569, "num_tokens": 808605239.0, "step": 21191 }, { "epoch": 2.6958402238900905, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.391742706298828, "learning_rate": 1e-06, "loss": 0.4643, "mean_token_accuracy": 0.8525109887123108, "num_tokens": 808643227.0, "step": 21192 }, { "epoch": 2.695967434168681, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.410112380981445, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8693387508392334, "num_tokens": 808683855.0, "step": 21193 }, { "epoch": 2.696094644447271, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.364703178405762, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8808832764625549, "num_tokens": 808723467.0, "step": 21194 }, { "epoch": 2.696221854725862, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.44001579284668, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8638389110565186, "num_tokens": 808759805.0, "step": 21195 }, { "epoch": 2.696349065004452, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.390965461730957, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8577214479446411, "num_tokens": 808802024.0, "step": 21196 }, { "epoch": 2.696476275283043, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.368249893188477, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8667943477630615, "num_tokens": 808837086.0, "step": 21197 }, { "epoch": 2.696603485561633, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.426140785217285, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8735809922218323, "num_tokens": 808875261.0, "step": 21198 }, { "epoch": 2.6967306958402237, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.357377052307129, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8808008432388306, "num_tokens": 808913044.0, "step": 21199 }, { "epoch": 2.6968579061188143, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.383271217346191, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8823315501213074, "num_tokens": 808952374.0, "step": 21200 }, { "epoch": 2.696985116397405, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.317991256713867, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8758046627044678, "num_tokens": 808994397.0, "step": 21201 }, { "epoch": 2.6971123266759953, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.389660835266113, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.876975417137146, "num_tokens": 809036401.0, "step": 21202 }, { "epoch": 2.697239536954586, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.338371276855469, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8800315260887146, "num_tokens": 809073073.0, "step": 21203 }, { "epoch": 2.6973667472331764, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.35378646850586, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.87399822473526, "num_tokens": 809117348.0, "step": 21204 }, { "epoch": 2.697493957511767, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.312746047973633, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8879865407943726, "num_tokens": 809151357.0, "step": 21205 }, { "epoch": 2.6976211677903574, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.345582008361816, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8689388632774353, "num_tokens": 809189000.0, "step": 21206 }, { "epoch": 2.697748378068948, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.3338041305542, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8782299757003784, "num_tokens": 809227380.0, "step": 21207 }, { "epoch": 2.6978755883475385, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.359045028686523, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.874800443649292, "num_tokens": 809265131.0, "step": 21208 }, { "epoch": 2.698002798626129, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.41180419921875, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8743877410888672, "num_tokens": 809297190.0, "step": 21209 }, { "epoch": 2.6981300089047195, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.325723648071289, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8843122720718384, "num_tokens": 809336774.0, "step": 21210 }, { "epoch": 2.69825721918331, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4020414352417, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.858116090297699, "num_tokens": 809376612.0, "step": 21211 }, { "epoch": 2.6983844294619006, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.226646423339844, "learning_rate": 1e-06, "loss": 0.3231, "mean_token_accuracy": 0.8974049091339111, "num_tokens": 809416058.0, "step": 21212 }, { "epoch": 2.698511639740491, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.380437850952148, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8846602439880371, "num_tokens": 809449580.0, "step": 21213 }, { "epoch": 2.6986388500190817, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.308317184448242, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8778399229049683, "num_tokens": 809484983.0, "step": 21214 }, { "epoch": 2.698766060297672, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.31126880645752, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8781745433807373, "num_tokens": 809521086.0, "step": 21215 }, { "epoch": 2.6988932705762627, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330923080444336, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.8724392652511597, "num_tokens": 809562475.0, "step": 21216 }, { "epoch": 2.699020480854853, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45028018951416, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8610454797744751, "num_tokens": 809596686.0, "step": 21217 }, { "epoch": 2.6991476911334438, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.28166389465332, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8758206367492676, "num_tokens": 809636614.0, "step": 21218 }, { "epoch": 2.699274901412034, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.390989303588867, "learning_rate": 1e-06, "loss": 0.4363, "mean_token_accuracy": 0.8627357482910156, "num_tokens": 809674542.0, "step": 21219 }, { "epoch": 2.699402111690625, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.322467803955078, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8904558420181274, "num_tokens": 809716726.0, "step": 21220 }, { "epoch": 2.699529321969215, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.23210620880127, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8837274312973022, "num_tokens": 809756790.0, "step": 21221 }, { "epoch": 2.699656532247806, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.384430885314941, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8720918297767639, "num_tokens": 809794616.0, "step": 21222 }, { "epoch": 2.699783742526396, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.273787498474121, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8811125159263611, "num_tokens": 809836662.0, "step": 21223 }, { "epoch": 2.6999109528049865, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.351930618286133, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8716498613357544, "num_tokens": 809873812.0, "step": 21224 }, { "epoch": 2.700038163083577, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.33897590637207, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8746323585510254, "num_tokens": 809913139.0, "step": 21225 }, { "epoch": 2.7001653733621676, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.373801231384277, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8761023283004761, "num_tokens": 809947741.0, "step": 21226 }, { "epoch": 2.700292583640758, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.353842735290527, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8675283193588257, "num_tokens": 809987120.0, "step": 21227 }, { "epoch": 2.7004197939193486, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.341484069824219, "learning_rate": 1e-06, "loss": 0.4205, "mean_token_accuracy": 0.8694815635681152, "num_tokens": 810026972.0, "step": 21228 }, { "epoch": 2.700547004197939, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.325818061828613, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8743410110473633, "num_tokens": 810062767.0, "step": 21229 }, { "epoch": 2.7006742144765297, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.281004905700684, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8826147317886353, "num_tokens": 810096930.0, "step": 21230 }, { "epoch": 2.70080142475512, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.362664222717285, "learning_rate": 1e-06, "loss": 0.4404, "mean_token_accuracy": 0.8615976572036743, "num_tokens": 810134600.0, "step": 21231 }, { "epoch": 2.7009286350337107, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.286842346191406, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.875582218170166, "num_tokens": 810169535.0, "step": 21232 }, { "epoch": 2.7010558453123013, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.366941452026367, "learning_rate": 1e-06, "loss": 0.4303, "mean_token_accuracy": 0.8649778366088867, "num_tokens": 810205288.0, "step": 21233 }, { "epoch": 2.701183055590892, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.341667175292969, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8724054098129272, "num_tokens": 810242868.0, "step": 21234 }, { "epoch": 2.7013102658694823, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.355476379394531, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8628726005554199, "num_tokens": 810286539.0, "step": 21235 }, { "epoch": 2.701437476148073, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.332094192504883, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8777588605880737, "num_tokens": 810331598.0, "step": 21236 }, { "epoch": 2.7015646864266634, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.364710807800293, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8824440240859985, "num_tokens": 810369501.0, "step": 21237 }, { "epoch": 2.701691896705254, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.305354118347168, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8709537982940674, "num_tokens": 810410727.0, "step": 21238 }, { "epoch": 2.7018191069838444, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.388628959655762, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8665100336074829, "num_tokens": 810450315.0, "step": 21239 }, { "epoch": 2.701946317262435, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.35216236114502, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.8671971559524536, "num_tokens": 810493972.0, "step": 21240 }, { "epoch": 2.7020735275410255, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.328600883483887, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8855854868888855, "num_tokens": 810526038.0, "step": 21241 }, { "epoch": 2.7022007378196156, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.40312385559082, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8788536787033081, "num_tokens": 810560772.0, "step": 21242 }, { "epoch": 2.7023279480982065, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.32541561126709, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8820586800575256, "num_tokens": 810598322.0, "step": 21243 }, { "epoch": 2.7024551583767966, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.357476234436035, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8936166167259216, "num_tokens": 810638652.0, "step": 21244 }, { "epoch": 2.7025823686553876, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.35601806640625, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8732635974884033, "num_tokens": 810680318.0, "step": 21245 }, { "epoch": 2.7027095789339777, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.366812705993652, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8784193992614746, "num_tokens": 810715001.0, "step": 21246 }, { "epoch": 2.7028367892125686, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.338391304016113, "learning_rate": 1e-06, "loss": 0.4374, "mean_token_accuracy": 0.8674042224884033, "num_tokens": 810748897.0, "step": 21247 }, { "epoch": 2.7029639994911587, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.384664535522461, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8686907291412354, "num_tokens": 810788954.0, "step": 21248 }, { "epoch": 2.7030912097697493, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.338937759399414, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.876970648765564, "num_tokens": 810828897.0, "step": 21249 }, { "epoch": 2.70321842004834, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.329493522644043, "learning_rate": 1e-06, "loss": 0.3438, "mean_token_accuracy": 0.8916076421737671, "num_tokens": 810866698.0, "step": 21250 }, { "epoch": 2.7033456303269303, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.341377258300781, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8670378923416138, "num_tokens": 810903140.0, "step": 21251 }, { "epoch": 2.703472840605521, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.35253620147705, "learning_rate": 1e-06, "loss": 0.3873, "mean_token_accuracy": 0.8791184425354004, "num_tokens": 810940664.0, "step": 21252 }, { "epoch": 2.7036000508841114, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.352272987365723, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8720009326934814, "num_tokens": 810981818.0, "step": 21253 }, { "epoch": 2.703727261162702, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.307339668273926, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8736600875854492, "num_tokens": 811019897.0, "step": 21254 }, { "epoch": 2.7038544714412924, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.405534744262695, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8781875371932983, "num_tokens": 811056310.0, "step": 21255 }, { "epoch": 2.703981681719883, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.409122467041016, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8639271855354309, "num_tokens": 811091505.0, "step": 21256 }, { "epoch": 2.7041088919984735, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.391646385192871, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8850995302200317, "num_tokens": 811134861.0, "step": 21257 }, { "epoch": 2.704236102277064, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.336360931396484, "learning_rate": 1e-06, "loss": 0.4197, "mean_token_accuracy": 0.8662751317024231, "num_tokens": 811170001.0, "step": 21258 }, { "epoch": 2.7043633125556545, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.380561828613281, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.878788411617279, "num_tokens": 811204203.0, "step": 21259 }, { "epoch": 2.704490522834245, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.342045783996582, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8790496587753296, "num_tokens": 811244598.0, "step": 21260 }, { "epoch": 2.7046177331128356, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.328564643859863, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8845367431640625, "num_tokens": 811283352.0, "step": 21261 }, { "epoch": 2.704744943391426, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.374004364013672, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8626663088798523, "num_tokens": 811317605.0, "step": 21262 }, { "epoch": 2.7048721536700167, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.38642406463623, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8914679884910583, "num_tokens": 811352117.0, "step": 21263 }, { "epoch": 2.704999363948607, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330519676208496, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8669651746749878, "num_tokens": 811386118.0, "step": 21264 }, { "epoch": 2.7051265742271977, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.356210708618164, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8849684596061707, "num_tokens": 811424479.0, "step": 21265 }, { "epoch": 2.7052537845057882, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.338642120361328, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8621368408203125, "num_tokens": 811464683.0, "step": 21266 }, { "epoch": 2.7053809947843783, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.349209785461426, "learning_rate": 1e-06, "loss": 0.3548, "mean_token_accuracy": 0.8894660472869873, "num_tokens": 811504801.0, "step": 21267 }, { "epoch": 2.7055082050629693, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.41220760345459, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8671747446060181, "num_tokens": 811546278.0, "step": 21268 }, { "epoch": 2.7056354153415594, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.326870918273926, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8614343404769897, "num_tokens": 811587019.0, "step": 21269 }, { "epoch": 2.7057626256201504, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.3632230758667, "learning_rate": 1e-06, "loss": 0.4658, "mean_token_accuracy": 0.8500412702560425, "num_tokens": 811619843.0, "step": 21270 }, { "epoch": 2.7058898358987404, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.352022171020508, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8799933195114136, "num_tokens": 811655797.0, "step": 21271 }, { "epoch": 2.706017046177331, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.336984634399414, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8663418292999268, "num_tokens": 811691626.0, "step": 21272 }, { "epoch": 2.7061442564559215, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.366439819335938, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8764629364013672, "num_tokens": 811726046.0, "step": 21273 }, { "epoch": 2.706271466734512, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.327457427978516, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8645122051239014, "num_tokens": 811765686.0, "step": 21274 }, { "epoch": 2.7063986770131025, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.355828285217285, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8813531398773193, "num_tokens": 811801143.0, "step": 21275 }, { "epoch": 2.706525887291693, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.378288269042969, "learning_rate": 1e-06, "loss": 0.3697, "mean_token_accuracy": 0.8814541101455688, "num_tokens": 811834807.0, "step": 21276 }, { "epoch": 2.7066530975702836, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.330982208251953, "learning_rate": 1e-06, "loss": 0.3442, "mean_token_accuracy": 0.8932982087135315, "num_tokens": 811865676.0, "step": 21277 }, { "epoch": 2.706780307848874, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.328729629516602, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8724163770675659, "num_tokens": 811901259.0, "step": 21278 }, { "epoch": 2.7069075181274647, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.354413986206055, "learning_rate": 1e-06, "loss": 0.4495, "mean_token_accuracy": 0.8608917593955994, "num_tokens": 811937768.0, "step": 21279 }, { "epoch": 2.707034728406055, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.350552558898926, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.874672532081604, "num_tokens": 811967805.0, "step": 21280 }, { "epoch": 2.7071619386846457, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.318639755249023, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.870835542678833, "num_tokens": 812007468.0, "step": 21281 }, { "epoch": 2.7072891489632362, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.330794334411621, "learning_rate": 1e-06, "loss": 0.3892, "mean_token_accuracy": 0.875998318195343, "num_tokens": 812048127.0, "step": 21282 }, { "epoch": 2.7074163592418268, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.341470718383789, "learning_rate": 1e-06, "loss": 0.3882, "mean_token_accuracy": 0.875546932220459, "num_tokens": 812083904.0, "step": 21283 }, { "epoch": 2.7075435695204173, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.321104049682617, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8797078728675842, "num_tokens": 812119643.0, "step": 21284 }, { "epoch": 2.707670779799008, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.348129272460938, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.873163104057312, "num_tokens": 812155338.0, "step": 21285 }, { "epoch": 2.7077979900775984, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.375476837158203, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8740279674530029, "num_tokens": 812193075.0, "step": 21286 }, { "epoch": 2.707925200356189, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.37479019165039, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8837997913360596, "num_tokens": 812228163.0, "step": 21287 }, { "epoch": 2.7080524106347794, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.454080581665039, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8690096139907837, "num_tokens": 812268565.0, "step": 21288 }, { "epoch": 2.70817962091337, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.348111152648926, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8797661662101746, "num_tokens": 812307292.0, "step": 21289 }, { "epoch": 2.7083068311919605, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.360112190246582, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8845075964927673, "num_tokens": 812346180.0, "step": 21290 }, { "epoch": 2.708434041470551, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.352166175842285, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8722703456878662, "num_tokens": 812376033.0, "step": 21291 }, { "epoch": 2.708561251749141, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.417831420898438, "learning_rate": 1e-06, "loss": 0.3467, "mean_token_accuracy": 0.8924225568771362, "num_tokens": 812407925.0, "step": 21292 }, { "epoch": 2.708688462027732, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.38964557647705, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8800565004348755, "num_tokens": 812452354.0, "step": 21293 }, { "epoch": 2.708815672306322, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.417244911193848, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8742860555648804, "num_tokens": 812490015.0, "step": 21294 }, { "epoch": 2.708942882584913, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.376601219177246, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8712370991706848, "num_tokens": 812529919.0, "step": 21295 }, { "epoch": 2.709070092863503, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.409687042236328, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8872320652008057, "num_tokens": 812568541.0, "step": 21296 }, { "epoch": 2.7091973031420937, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.348841667175293, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8670938014984131, "num_tokens": 812612098.0, "step": 21297 }, { "epoch": 2.7093245134206843, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.406607627868652, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8711903095245361, "num_tokens": 812648843.0, "step": 21298 }, { "epoch": 2.709451723699275, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.293554306030273, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.875311017036438, "num_tokens": 812692080.0, "step": 21299 }, { "epoch": 2.7095789339778653, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.448009490966797, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8881639242172241, "num_tokens": 812729441.0, "step": 21300 }, { "epoch": 2.709706144256456, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.363273620605469, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8677895069122314, "num_tokens": 812768908.0, "step": 21301 }, { "epoch": 2.7098333545350464, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.352373123168945, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.874025285243988, "num_tokens": 812808384.0, "step": 21302 }, { "epoch": 2.709960564813637, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.395331382751465, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.866356372833252, "num_tokens": 812847658.0, "step": 21303 }, { "epoch": 2.7100877750922274, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.326480865478516, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8753730654716492, "num_tokens": 812889394.0, "step": 21304 }, { "epoch": 2.710214985370818, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.3696870803833, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8629271984100342, "num_tokens": 812924202.0, "step": 21305 }, { "epoch": 2.7103421956494085, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.401623725891113, "learning_rate": 1e-06, "loss": 0.3354, "mean_token_accuracy": 0.8938122987747192, "num_tokens": 812954391.0, "step": 21306 }, { "epoch": 2.710469405927999, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.392860412597656, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.8666403293609619, "num_tokens": 812989498.0, "step": 21307 }, { "epoch": 2.7105966162065895, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.37321949005127, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8811360597610474, "num_tokens": 813030968.0, "step": 21308 }, { "epoch": 2.71072382648518, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.393898010253906, "learning_rate": 1e-06, "loss": 0.3295, "mean_token_accuracy": 0.896453857421875, "num_tokens": 813065843.0, "step": 21309 }, { "epoch": 2.7108510367637706, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.412945747375488, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8805681467056274, "num_tokens": 813107410.0, "step": 21310 }, { "epoch": 2.710978247042361, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.368341445922852, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8812085390090942, "num_tokens": 813155959.0, "step": 21311 }, { "epoch": 2.7111054573209517, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.411791801452637, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8814237117767334, "num_tokens": 813194133.0, "step": 21312 }, { "epoch": 2.711232667599542, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.390850067138672, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8623670339584351, "num_tokens": 813231015.0, "step": 21313 }, { "epoch": 2.7113598778781327, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.405057907104492, "learning_rate": 1e-06, "loss": 0.3425, "mean_token_accuracy": 0.890513002872467, "num_tokens": 813271497.0, "step": 21314 }, { "epoch": 2.711487088156723, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.380487442016602, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8763487339019775, "num_tokens": 813305758.0, "step": 21315 }, { "epoch": 2.7116142984353138, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.413151741027832, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8735471367835999, "num_tokens": 813345958.0, "step": 21316 }, { "epoch": 2.711741508713904, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417033195495605, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.878359317779541, "num_tokens": 813381630.0, "step": 21317 }, { "epoch": 2.711868718992495, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.382491111755371, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8752303123474121, "num_tokens": 813421320.0, "step": 21318 }, { "epoch": 2.711995929271085, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.402190208435059, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8685722947120667, "num_tokens": 813458323.0, "step": 21319 }, { "epoch": 2.712123139549676, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.403165817260742, "learning_rate": 1e-06, "loss": 0.4575, "mean_token_accuracy": 0.857345461845398, "num_tokens": 813501436.0, "step": 21320 }, { "epoch": 2.712250349828266, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.297590255737305, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8757532238960266, "num_tokens": 813545942.0, "step": 21321 }, { "epoch": 2.7123775601068565, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.409781455993652, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.870450496673584, "num_tokens": 813582202.0, "step": 21322 }, { "epoch": 2.712504770385447, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.360027313232422, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8600070476531982, "num_tokens": 813621711.0, "step": 21323 }, { "epoch": 2.7126319806640375, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.321359634399414, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.869284987449646, "num_tokens": 813665404.0, "step": 21324 }, { "epoch": 2.712759190942628, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.368963241577148, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8816555738449097, "num_tokens": 813697470.0, "step": 21325 }, { "epoch": 2.7128864012212186, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.356849670410156, "learning_rate": 1e-06, "loss": 0.3743, "mean_token_accuracy": 0.8818235993385315, "num_tokens": 813733308.0, "step": 21326 }, { "epoch": 2.713013611499809, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.3577880859375, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8608002662658691, "num_tokens": 813770881.0, "step": 21327 }, { "epoch": 2.7131408217783997, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.416958808898926, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8814104199409485, "num_tokens": 813806495.0, "step": 21328 }, { "epoch": 2.71326803205699, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.322104454040527, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.880052387714386, "num_tokens": 813842408.0, "step": 21329 }, { "epoch": 2.7133952423355807, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.378339767456055, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8705424070358276, "num_tokens": 813883041.0, "step": 21330 }, { "epoch": 2.7135224526141712, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.363709449768066, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8711071014404297, "num_tokens": 813922901.0, "step": 21331 }, { "epoch": 2.7136496628927618, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.375271797180176, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8777079582214355, "num_tokens": 813961948.0, "step": 21332 }, { "epoch": 2.7137768731713523, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.346855163574219, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8890500068664551, "num_tokens": 813999194.0, "step": 21333 }, { "epoch": 2.713904083449943, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.368739128112793, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8640411496162415, "num_tokens": 814038986.0, "step": 21334 }, { "epoch": 2.7140312937285334, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.374543190002441, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8879790306091309, "num_tokens": 814074351.0, "step": 21335 }, { "epoch": 2.714158504007124, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.331650733947754, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8740968704223633, "num_tokens": 814109255.0, "step": 21336 }, { "epoch": 2.7142857142857144, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.376768112182617, "learning_rate": 1e-06, "loss": 0.4113, "mean_token_accuracy": 0.8707982301712036, "num_tokens": 814146601.0, "step": 21337 }, { "epoch": 2.714412924564305, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.379867553710938, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8795846104621887, "num_tokens": 814181299.0, "step": 21338 }, { "epoch": 2.7145401348428955, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.248876571655273, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8831831216812134, "num_tokens": 814218178.0, "step": 21339 }, { "epoch": 2.7146673451214856, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.345725059509277, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8767789602279663, "num_tokens": 814254161.0, "step": 21340 }, { "epoch": 2.7147945554000765, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.359476089477539, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8752299547195435, "num_tokens": 814290164.0, "step": 21341 }, { "epoch": 2.7149217656786666, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.344554901123047, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8548601865768433, "num_tokens": 814334459.0, "step": 21342 }, { "epoch": 2.7150489759572576, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.351449012756348, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8667611479759216, "num_tokens": 814374291.0, "step": 21343 }, { "epoch": 2.7151761862358477, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.439382553100586, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8700759410858154, "num_tokens": 814413196.0, "step": 21344 }, { "epoch": 2.7153033965144386, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.32709789276123, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8738806247711182, "num_tokens": 814455312.0, "step": 21345 }, { "epoch": 2.7154306067930287, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.331429481506348, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8704123497009277, "num_tokens": 814496586.0, "step": 21346 }, { "epoch": 2.7155578170716193, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.38628101348877, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8800476789474487, "num_tokens": 814531566.0, "step": 21347 }, { "epoch": 2.71568502735021, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.400689125061035, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8815828561782837, "num_tokens": 814566168.0, "step": 21348 }, { "epoch": 2.7158122376288003, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.364429473876953, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8746136426925659, "num_tokens": 814604927.0, "step": 21349 }, { "epoch": 2.715939447907391, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.392687797546387, "learning_rate": 1e-06, "loss": 0.3536, "mean_token_accuracy": 0.8885325193405151, "num_tokens": 814641401.0, "step": 21350 }, { "epoch": 2.7160666581859814, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.353523254394531, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8812849521636963, "num_tokens": 814674252.0, "step": 21351 }, { "epoch": 2.716193868464572, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.439935684204102, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8742060661315918, "num_tokens": 814704965.0, "step": 21352 }, { "epoch": 2.7163210787431624, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.383649826049805, "learning_rate": 1e-06, "loss": 0.4805, "mean_token_accuracy": 0.8469483852386475, "num_tokens": 814743740.0, "step": 21353 }, { "epoch": 2.716448289021753, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.35903263092041, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8807259798049927, "num_tokens": 814784598.0, "step": 21354 }, { "epoch": 2.7165754993003435, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.350438117980957, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8686638474464417, "num_tokens": 814821437.0, "step": 21355 }, { "epoch": 2.716702709578934, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.418109893798828, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8751577138900757, "num_tokens": 814861638.0, "step": 21356 }, { "epoch": 2.7168299198575245, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.335867881774902, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8769941926002502, "num_tokens": 814893799.0, "step": 21357 }, { "epoch": 2.716957130136115, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.390082359313965, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8818036317825317, "num_tokens": 814930726.0, "step": 21358 }, { "epoch": 2.7170843404147056, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.412654876708984, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8740711212158203, "num_tokens": 814970041.0, "step": 21359 }, { "epoch": 2.717211550693296, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.328848838806152, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8785744905471802, "num_tokens": 815008587.0, "step": 21360 }, { "epoch": 2.7173387609718866, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.36899471282959, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.862345814704895, "num_tokens": 815047659.0, "step": 21361 }, { "epoch": 2.717465971250477, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.338441848754883, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8751332759857178, "num_tokens": 815082631.0, "step": 21362 }, { "epoch": 2.7175931815290677, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.385193824768066, "learning_rate": 1e-06, "loss": 0.339, "mean_token_accuracy": 0.8925328254699707, "num_tokens": 815117568.0, "step": 21363 }, { "epoch": 2.7177203918076582, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.377877235412598, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8755332231521606, "num_tokens": 815155019.0, "step": 21364 }, { "epoch": 2.7178476020862483, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.301857948303223, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8748307228088379, "num_tokens": 815195386.0, "step": 21365 }, { "epoch": 2.7179748123648393, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.409131050109863, "learning_rate": 1e-06, "loss": 0.388, "mean_token_accuracy": 0.8793131709098816, "num_tokens": 815228304.0, "step": 21366 }, { "epoch": 2.7181020226434294, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.380711555480957, "learning_rate": 1e-06, "loss": 0.3328, "mean_token_accuracy": 0.8938170671463013, "num_tokens": 815260111.0, "step": 21367 }, { "epoch": 2.7182292329220203, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.401315689086914, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8742705583572388, "num_tokens": 815297272.0, "step": 21368 }, { "epoch": 2.7183564432006104, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.358085632324219, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8741781115531921, "num_tokens": 815336016.0, "step": 21369 }, { "epoch": 2.718483653479201, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.45871353149414, "learning_rate": 1e-06, "loss": 0.4733, "mean_token_accuracy": 0.8493014574050903, "num_tokens": 815373940.0, "step": 21370 }, { "epoch": 2.7186108637577915, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.397536277770996, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8645510077476501, "num_tokens": 815410836.0, "step": 21371 }, { "epoch": 2.718738074036382, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.356903076171875, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8801416158676147, "num_tokens": 815449577.0, "step": 21372 }, { "epoch": 2.7188652843149725, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.441956520080566, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8775415420532227, "num_tokens": 815485869.0, "step": 21373 }, { "epoch": 2.718992494593563, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.352380752563477, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8571873307228088, "num_tokens": 815523614.0, "step": 21374 }, { "epoch": 2.7191197048721536, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.41185188293457, "learning_rate": 1e-06, "loss": 0.4543, "mean_token_accuracy": 0.8546572923660278, "num_tokens": 815561835.0, "step": 21375 }, { "epoch": 2.719246915150744, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.2986421585083, "learning_rate": 1e-06, "loss": 0.4421, "mean_token_accuracy": 0.8600640296936035, "num_tokens": 815605596.0, "step": 21376 }, { "epoch": 2.7193741254293347, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.368939399719238, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8690546154975891, "num_tokens": 815638720.0, "step": 21377 }, { "epoch": 2.719501335707925, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.334344863891602, "learning_rate": 1e-06, "loss": 0.3515, "mean_token_accuracy": 0.8891475796699524, "num_tokens": 815673996.0, "step": 21378 }, { "epoch": 2.7196285459865157, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.375005722045898, "learning_rate": 1e-06, "loss": 0.4381, "mean_token_accuracy": 0.8629227876663208, "num_tokens": 815716571.0, "step": 21379 }, { "epoch": 2.7197557562651062, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.356634140014648, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8816332221031189, "num_tokens": 815748973.0, "step": 21380 }, { "epoch": 2.7198829665436968, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.323251724243164, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8769736886024475, "num_tokens": 815794754.0, "step": 21381 }, { "epoch": 2.7200101768222873, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.376689910888672, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8780708312988281, "num_tokens": 815831051.0, "step": 21382 }, { "epoch": 2.720137387100878, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.39172649383545, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8769868016242981, "num_tokens": 815874291.0, "step": 21383 }, { "epoch": 2.7202645973794684, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.390030860900879, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8718099594116211, "num_tokens": 815912484.0, "step": 21384 }, { "epoch": 2.720391807658059, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.368675231933594, "learning_rate": 1e-06, "loss": 0.4522, "mean_token_accuracy": 0.8586486577987671, "num_tokens": 815948626.0, "step": 21385 }, { "epoch": 2.7205190179366494, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.39975643157959, "learning_rate": 1e-06, "loss": 0.3592, "mean_token_accuracy": 0.8841221928596497, "num_tokens": 815983154.0, "step": 21386 }, { "epoch": 2.72064622821524, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.395944595336914, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8772156238555908, "num_tokens": 816017524.0, "step": 21387 }, { "epoch": 2.7207734384938305, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.363677024841309, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8699272871017456, "num_tokens": 816055204.0, "step": 21388 }, { "epoch": 2.720900648772421, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417417526245117, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8825201988220215, "num_tokens": 816086671.0, "step": 21389 }, { "epoch": 2.721027859051011, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.467257499694824, "learning_rate": 1e-06, "loss": 0.4734, "mean_token_accuracy": 0.8559619784355164, "num_tokens": 816114173.0, "step": 21390 }, { "epoch": 2.721155069329602, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.374138832092285, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.887316107749939, "num_tokens": 816156434.0, "step": 21391 }, { "epoch": 2.721282279608192, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.372750282287598, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.870278000831604, "num_tokens": 816191037.0, "step": 21392 }, { "epoch": 2.721409489886783, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.384553909301758, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8878466486930847, "num_tokens": 816227031.0, "step": 21393 }, { "epoch": 2.721536700165373, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.375210762023926, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.878581702709198, "num_tokens": 816264308.0, "step": 21394 }, { "epoch": 2.7216639104439637, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.409751892089844, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.873325765132904, "num_tokens": 816300955.0, "step": 21395 }, { "epoch": 2.7217911207225542, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.387585639953613, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8759762644767761, "num_tokens": 816332881.0, "step": 21396 }, { "epoch": 2.7219183310011448, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.357634544372559, "learning_rate": 1e-06, "loss": 0.3412, "mean_token_accuracy": 0.8936752080917358, "num_tokens": 816367446.0, "step": 21397 }, { "epoch": 2.7220455412797353, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.329150199890137, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.859066903591156, "num_tokens": 816407286.0, "step": 21398 }, { "epoch": 2.722172751558326, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.387969970703125, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8834159970283508, "num_tokens": 816445277.0, "step": 21399 }, { "epoch": 2.7222999618369164, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.375633239746094, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8796002864837646, "num_tokens": 816481638.0, "step": 21400 }, { "epoch": 2.722427172115507, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.3286714553833, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8766520023345947, "num_tokens": 816521386.0, "step": 21401 }, { "epoch": 2.7225543823940974, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.448594093322754, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8863750696182251, "num_tokens": 816554760.0, "step": 21402 }, { "epoch": 2.722681592672688, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.309499740600586, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8834457993507385, "num_tokens": 816595195.0, "step": 21403 }, { "epoch": 2.7228088029512785, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.465263366699219, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8770554065704346, "num_tokens": 816633046.0, "step": 21404 }, { "epoch": 2.722936013229869, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.295073509216309, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.889258861541748, "num_tokens": 816675283.0, "step": 21405 }, { "epoch": 2.7230632235084595, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.388453483581543, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8657447695732117, "num_tokens": 816716499.0, "step": 21406 }, { "epoch": 2.72319043378705, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.423550605773926, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8688144683837891, "num_tokens": 816761825.0, "step": 21407 }, { "epoch": 2.7233176440656406, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.37483024597168, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8817386031150818, "num_tokens": 816792080.0, "step": 21408 }, { "epoch": 2.723444854344231, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.422287940979004, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8655663132667542, "num_tokens": 816832121.0, "step": 21409 }, { "epoch": 2.7235720646228216, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.351189613342285, "learning_rate": 1e-06, "loss": 0.4234, "mean_token_accuracy": 0.8685426712036133, "num_tokens": 816874032.0, "step": 21410 }, { "epoch": 2.723699274901412, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.417901992797852, "learning_rate": 1e-06, "loss": 0.453, "mean_token_accuracy": 0.8573011159896851, "num_tokens": 816920471.0, "step": 21411 }, { "epoch": 2.7238264851800027, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.316556930541992, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8755617737770081, "num_tokens": 816958042.0, "step": 21412 }, { "epoch": 2.723953695458593, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.481846809387207, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8659656047821045, "num_tokens": 816992158.0, "step": 21413 }, { "epoch": 2.7240809057371838, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.404158592224121, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8643914461135864, "num_tokens": 817032552.0, "step": 21414 }, { "epoch": 2.724208116015774, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.377203941345215, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.872714102268219, "num_tokens": 817072240.0, "step": 21415 }, { "epoch": 2.724335326294365, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.318693161010742, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8653463125228882, "num_tokens": 817104741.0, "step": 21416 }, { "epoch": 2.724462536572955, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.450530052185059, "learning_rate": 1e-06, "loss": 0.4407, "mean_token_accuracy": 0.8616876006126404, "num_tokens": 817143876.0, "step": 21417 }, { "epoch": 2.724589746851546, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.361140251159668, "learning_rate": 1e-06, "loss": 0.3236, "mean_token_accuracy": 0.8976027369499207, "num_tokens": 817183647.0, "step": 21418 }, { "epoch": 2.724716957130136, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4036865234375, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8696090579032898, "num_tokens": 817214520.0, "step": 21419 }, { "epoch": 2.7248441674087265, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.354196548461914, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8651020526885986, "num_tokens": 817256936.0, "step": 21420 }, { "epoch": 2.724971377687317, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.358546257019043, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8754446506500244, "num_tokens": 817292840.0, "step": 21421 }, { "epoch": 2.7250985879659075, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.427978515625, "learning_rate": 1e-06, "loss": 0.4273, "mean_token_accuracy": 0.8709510564804077, "num_tokens": 817332872.0, "step": 21422 }, { "epoch": 2.725225798244498, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42650032043457, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8830480575561523, "num_tokens": 817367668.0, "step": 21423 }, { "epoch": 2.7253530085230886, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.348001480102539, "learning_rate": 1e-06, "loss": 0.3425, "mean_token_accuracy": 0.89333575963974, "num_tokens": 817407726.0, "step": 21424 }, { "epoch": 2.725480218801679, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42088508605957, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.8850221633911133, "num_tokens": 817441003.0, "step": 21425 }, { "epoch": 2.7256074290802697, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.333386421203613, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8798061609268188, "num_tokens": 817482587.0, "step": 21426 }, { "epoch": 2.72573463935886, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.389981269836426, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8797121047973633, "num_tokens": 817519727.0, "step": 21427 }, { "epoch": 2.7258618496374507, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.340470314025879, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8723771572113037, "num_tokens": 817553809.0, "step": 21428 }, { "epoch": 2.7259890599160412, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.317388534545898, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.858659029006958, "num_tokens": 817588893.0, "step": 21429 }, { "epoch": 2.7261162701946318, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.40059757232666, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8812255263328552, "num_tokens": 817627789.0, "step": 21430 }, { "epoch": 2.7262434804732223, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.368672370910645, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8654396533966064, "num_tokens": 817670048.0, "step": 21431 }, { "epoch": 2.726370690751813, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.358308792114258, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8799569010734558, "num_tokens": 817707108.0, "step": 21432 }, { "epoch": 2.7264979010304033, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.42061996459961, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8733952045440674, "num_tokens": 817748126.0, "step": 21433 }, { "epoch": 2.726625111308994, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.310141563415527, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8778769969940186, "num_tokens": 817781578.0, "step": 21434 }, { "epoch": 2.7267523215875844, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.478992462158203, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8819376826286316, "num_tokens": 817822143.0, "step": 21435 }, { "epoch": 2.726879531866175, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.33450984954834, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8857758045196533, "num_tokens": 817858456.0, "step": 21436 }, { "epoch": 2.7270067421447655, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.385702133178711, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.8812874555587769, "num_tokens": 817892863.0, "step": 21437 }, { "epoch": 2.7271339524233555, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.358028411865234, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8791316151618958, "num_tokens": 817931120.0, "step": 21438 }, { "epoch": 2.7272611627019465, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.375786781311035, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8784640431404114, "num_tokens": 817963076.0, "step": 21439 }, { "epoch": 2.7273883729805366, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.459179878234863, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8761743307113647, "num_tokens": 818003844.0, "step": 21440 }, { "epoch": 2.7275155832591276, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.401823997497559, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.878181517124176, "num_tokens": 818048533.0, "step": 21441 }, { "epoch": 2.7276427935377177, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.430685997009277, "learning_rate": 1e-06, "loss": 0.4214, "mean_token_accuracy": 0.8718831539154053, "num_tokens": 818083290.0, "step": 21442 }, { "epoch": 2.7277700038163086, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.44980525970459, "learning_rate": 1e-06, "loss": 0.365, "mean_token_accuracy": 0.8834916353225708, "num_tokens": 818120037.0, "step": 21443 }, { "epoch": 2.7278972140948987, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.394448280334473, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8754837512969971, "num_tokens": 818164122.0, "step": 21444 }, { "epoch": 2.7280244243734892, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.4472017288208, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8775215148925781, "num_tokens": 818198065.0, "step": 21445 }, { "epoch": 2.7281516346520798, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.44135570526123, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8767434358596802, "num_tokens": 818236843.0, "step": 21446 }, { "epoch": 2.7282788449306703, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.42600154876709, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8768233060836792, "num_tokens": 818271300.0, "step": 21447 }, { "epoch": 2.728406055209261, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.509966850280762, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.861253023147583, "num_tokens": 818311799.0, "step": 21448 }, { "epoch": 2.7285332654878514, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.385066032409668, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8804304599761963, "num_tokens": 818346662.0, "step": 21449 }, { "epoch": 2.728660475766442, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.475878715515137, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8795586824417114, "num_tokens": 818384465.0, "step": 21450 }, { "epoch": 2.7287876860450324, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.392192840576172, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8730130195617676, "num_tokens": 818425625.0, "step": 21451 }, { "epoch": 2.728914896323623, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.405760765075684, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8801432847976685, "num_tokens": 818458544.0, "step": 21452 }, { "epoch": 2.7290421066022135, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.352231979370117, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8705135583877563, "num_tokens": 818503507.0, "step": 21453 }, { "epoch": 2.729169316880804, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.39941120147705, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8815752267837524, "num_tokens": 818545411.0, "step": 21454 }, { "epoch": 2.7292965271593945, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.38033390045166, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8701865077018738, "num_tokens": 818582100.0, "step": 21455 }, { "epoch": 2.729423737437985, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.447993278503418, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8720481395721436, "num_tokens": 818615549.0, "step": 21456 }, { "epoch": 2.7295509477165756, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.371506690979004, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8793954849243164, "num_tokens": 818652476.0, "step": 21457 }, { "epoch": 2.729678157995166, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.387687683105469, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8790163397789001, "num_tokens": 818692435.0, "step": 21458 }, { "epoch": 2.7298053682737566, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.376562118530273, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8744806051254272, "num_tokens": 818733206.0, "step": 21459 }, { "epoch": 2.729932578552347, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.447211265563965, "learning_rate": 1e-06, "loss": 0.393, "mean_token_accuracy": 0.8794229030609131, "num_tokens": 818771717.0, "step": 21460 }, { "epoch": 2.7300597888309377, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.432690620422363, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8749818205833435, "num_tokens": 818806999.0, "step": 21461 }, { "epoch": 2.7301869991095282, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.37040901184082, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8806697130203247, "num_tokens": 818846226.0, "step": 21462 }, { "epoch": 2.7303142093881183, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.450519561767578, "learning_rate": 1e-06, "loss": 0.3696, "mean_token_accuracy": 0.8817355632781982, "num_tokens": 818885667.0, "step": 21463 }, { "epoch": 2.7304414196667093, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.339594841003418, "learning_rate": 1e-06, "loss": 0.3742, "mean_token_accuracy": 0.8806131482124329, "num_tokens": 818924009.0, "step": 21464 }, { "epoch": 2.7305686299452994, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.422248840332031, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8839127421379089, "num_tokens": 818962614.0, "step": 21465 }, { "epoch": 2.7306958402238903, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.43398380279541, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.877297043800354, "num_tokens": 819004486.0, "step": 21466 }, { "epoch": 2.7308230505024804, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.399249076843262, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8697288036346436, "num_tokens": 819048364.0, "step": 21467 }, { "epoch": 2.730950260781071, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.396628379821777, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8718324899673462, "num_tokens": 819086250.0, "step": 21468 }, { "epoch": 2.7310774710596615, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.439934730529785, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8860717415809631, "num_tokens": 819126705.0, "step": 21469 }, { "epoch": 2.731204681338252, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.408992767333984, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8770784139633179, "num_tokens": 819165967.0, "step": 21470 }, { "epoch": 2.7313318916168425, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.414094924926758, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8726056814193726, "num_tokens": 819208066.0, "step": 21471 }, { "epoch": 2.731459101895433, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.438520431518555, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8729844093322754, "num_tokens": 819246269.0, "step": 21472 }, { "epoch": 2.7315863121740236, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45132827758789, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8763437867164612, "num_tokens": 819286027.0, "step": 21473 }, { "epoch": 2.731713522452614, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.417780876159668, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8841662406921387, "num_tokens": 819319994.0, "step": 21474 }, { "epoch": 2.7318407327312046, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.438926696777344, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8716207146644592, "num_tokens": 819360113.0, "step": 21475 }, { "epoch": 2.731967943009795, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.424072265625, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8843951225280762, "num_tokens": 819394575.0, "step": 21476 }, { "epoch": 2.7320951532883857, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.404281616210938, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8703702688217163, "num_tokens": 819434545.0, "step": 21477 }, { "epoch": 2.7322223635669762, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.41560173034668, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8848162889480591, "num_tokens": 819470699.0, "step": 21478 }, { "epoch": 2.7323495738455668, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.456851959228516, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8701370358467102, "num_tokens": 819514447.0, "step": 21479 }, { "epoch": 2.7324767841241573, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.382980346679688, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8702151775360107, "num_tokens": 819553434.0, "step": 21480 }, { "epoch": 2.732603994402748, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.405893325805664, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8687398433685303, "num_tokens": 819591594.0, "step": 21481 }, { "epoch": 2.7327312046813383, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.368868827819824, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8711024522781372, "num_tokens": 819634530.0, "step": 21482 }, { "epoch": 2.732858414959929, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.428906440734863, "learning_rate": 1e-06, "loss": 0.4225, "mean_token_accuracy": 0.8662854433059692, "num_tokens": 819669358.0, "step": 21483 }, { "epoch": 2.7329856252385194, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417359352111816, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8752493262290955, "num_tokens": 819711336.0, "step": 21484 }, { "epoch": 2.73311283551711, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.402344703674316, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8665115833282471, "num_tokens": 819746973.0, "step": 21485 }, { "epoch": 2.7332400457957005, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.37637710571289, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8858051896095276, "num_tokens": 819780305.0, "step": 21486 }, { "epoch": 2.733367256074291, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.486305236816406, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8835985660552979, "num_tokens": 819817326.0, "step": 21487 }, { "epoch": 2.733494466352881, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.35645580291748, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8702483773231506, "num_tokens": 819857597.0, "step": 21488 }, { "epoch": 2.733621676631472, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.442026138305664, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8780753016471863, "num_tokens": 819893471.0, "step": 21489 }, { "epoch": 2.733748886910062, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.422589302062988, "learning_rate": 1e-06, "loss": 0.4541, "mean_token_accuracy": 0.8572934865951538, "num_tokens": 819934175.0, "step": 21490 }, { "epoch": 2.733876097188653, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.324764251708984, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8837184309959412, "num_tokens": 819971451.0, "step": 21491 }, { "epoch": 2.734003307467243, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.334152221679688, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8695710301399231, "num_tokens": 820011566.0, "step": 21492 }, { "epoch": 2.7341305177458337, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.362298965454102, "learning_rate": 1e-06, "loss": 0.4725, "mean_token_accuracy": 0.8529305458068848, "num_tokens": 820047944.0, "step": 21493 }, { "epoch": 2.7342577280244242, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.37433910369873, "learning_rate": 1e-06, "loss": 0.432, "mean_token_accuracy": 0.8631687164306641, "num_tokens": 820083416.0, "step": 21494 }, { "epoch": 2.7343849383030148, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.42398738861084, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8672720789909363, "num_tokens": 820122476.0, "step": 21495 }, { "epoch": 2.7345121485816053, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.381048202514648, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8806694149971008, "num_tokens": 820159866.0, "step": 21496 }, { "epoch": 2.734639358860196, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.355706214904785, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8755455017089844, "num_tokens": 820202710.0, "step": 21497 }, { "epoch": 2.7347665691387864, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.393584251403809, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8732984662055969, "num_tokens": 820239156.0, "step": 21498 }, { "epoch": 2.734893779417377, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.387421607971191, "learning_rate": 1e-06, "loss": 0.3337, "mean_token_accuracy": 0.8929685354232788, "num_tokens": 820274010.0, "step": 21499 }, { "epoch": 2.7350209896959674, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.38257884979248, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8725322484970093, "num_tokens": 820319171.0, "step": 21500 }, { "epoch": 2.735148199974558, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.374541282653809, "learning_rate": 1e-06, "loss": 0.4449, "mean_token_accuracy": 0.860787034034729, "num_tokens": 820355038.0, "step": 21501 }, { "epoch": 2.7352754102531485, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417454719543457, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8706182837486267, "num_tokens": 820390973.0, "step": 21502 }, { "epoch": 2.735402620531739, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.414824485778809, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8691028356552124, "num_tokens": 820424543.0, "step": 21503 }, { "epoch": 2.7355298308103295, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.429034233093262, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.8736077547073364, "num_tokens": 820462702.0, "step": 21504 }, { "epoch": 2.73565704108892, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.412359237670898, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.884655237197876, "num_tokens": 820498840.0, "step": 21505 }, { "epoch": 2.7357842513675106, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.43868637084961, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8687236905097961, "num_tokens": 820534083.0, "step": 21506 }, { "epoch": 2.735911461646101, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.444098472595215, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8781022429466248, "num_tokens": 820572479.0, "step": 21507 }, { "epoch": 2.7360386719246916, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.376945495605469, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8630644679069519, "num_tokens": 820610283.0, "step": 21508 }, { "epoch": 2.736165882203282, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.402687072753906, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8911944627761841, "num_tokens": 820647580.0, "step": 21509 }, { "epoch": 2.7362930924818727, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.395971298217773, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.88104248046875, "num_tokens": 820686774.0, "step": 21510 }, { "epoch": 2.7364203027604628, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.373849868774414, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8840901851654053, "num_tokens": 820724635.0, "step": 21511 }, { "epoch": 2.7365475130390537, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.425286293029785, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8814669847488403, "num_tokens": 820761467.0, "step": 21512 }, { "epoch": 2.736674723317644, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.347796440124512, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8620752692222595, "num_tokens": 820801001.0, "step": 21513 }, { "epoch": 2.736801933596235, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.41683292388916, "learning_rate": 1e-06, "loss": 0.4106, "mean_token_accuracy": 0.8700375556945801, "num_tokens": 820841028.0, "step": 21514 }, { "epoch": 2.736929143874825, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.419776916503906, "learning_rate": 1e-06, "loss": 0.3666, "mean_token_accuracy": 0.8830280303955078, "num_tokens": 820881613.0, "step": 21515 }, { "epoch": 2.737056354153416, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.387954711914062, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8821213245391846, "num_tokens": 820918205.0, "step": 21516 }, { "epoch": 2.737183564432006, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.402788162231445, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8683822154998779, "num_tokens": 820950751.0, "step": 21517 }, { "epoch": 2.7373107747105965, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.424060821533203, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8768519163131714, "num_tokens": 820987385.0, "step": 21518 }, { "epoch": 2.737437984989187, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.355942726135254, "learning_rate": 1e-06, "loss": 0.4219, "mean_token_accuracy": 0.8675857782363892, "num_tokens": 821026273.0, "step": 21519 }, { "epoch": 2.7375651952677775, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.364767074584961, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.866763174533844, "num_tokens": 821067978.0, "step": 21520 }, { "epoch": 2.737692405546368, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.383024215698242, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8804190754890442, "num_tokens": 821111925.0, "step": 21521 }, { "epoch": 2.7378196158249586, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.334078788757324, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8783785104751587, "num_tokens": 821151019.0, "step": 21522 }, { "epoch": 2.737946826103549, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.39254093170166, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8819930553436279, "num_tokens": 821185520.0, "step": 21523 }, { "epoch": 2.7380740363821396, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.401137351989746, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8773785829544067, "num_tokens": 821224273.0, "step": 21524 }, { "epoch": 2.73820124666073, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.410787582397461, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.880619466304779, "num_tokens": 821263541.0, "step": 21525 }, { "epoch": 2.7383284569393207, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.441758155822754, "learning_rate": 1e-06, "loss": 0.4329, "mean_token_accuracy": 0.863182783126831, "num_tokens": 821304053.0, "step": 21526 }, { "epoch": 2.7384556672179112, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.467350959777832, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.875519871711731, "num_tokens": 821339710.0, "step": 21527 }, { "epoch": 2.7385828774965018, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.404840469360352, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8746451139450073, "num_tokens": 821378377.0, "step": 21528 }, { "epoch": 2.7387100877750923, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.408443450927734, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8837481737136841, "num_tokens": 821417434.0, "step": 21529 }, { "epoch": 2.738837298053683, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.350358963012695, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8713634014129639, "num_tokens": 821464323.0, "step": 21530 }, { "epoch": 2.7389645083322733, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.466531753540039, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8723853826522827, "num_tokens": 821509891.0, "step": 21531 }, { "epoch": 2.739091718610864, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.42558479309082, "learning_rate": 1e-06, "loss": 0.439, "mean_token_accuracy": 0.8637646436691284, "num_tokens": 821557309.0, "step": 21532 }, { "epoch": 2.7392189288894544, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.364046096801758, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8704196214675903, "num_tokens": 821597751.0, "step": 21533 }, { "epoch": 2.739346139168045, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.435373306274414, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8871310949325562, "num_tokens": 821637961.0, "step": 21534 }, { "epoch": 2.7394733494466355, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.403016090393066, "learning_rate": 1e-06, "loss": 0.4471, "mean_token_accuracy": 0.8581541776657104, "num_tokens": 821684728.0, "step": 21535 }, { "epoch": 2.7396005597252255, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.400070190429688, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8801413774490356, "num_tokens": 821715359.0, "step": 21536 }, { "epoch": 2.7397277700038165, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.420402526855469, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8687305450439453, "num_tokens": 821755540.0, "step": 21537 }, { "epoch": 2.7398549802824066, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.3990478515625, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8757679462432861, "num_tokens": 821788506.0, "step": 21538 }, { "epoch": 2.7399821905609976, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.389811515808105, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.874053955078125, "num_tokens": 821826555.0, "step": 21539 }, { "epoch": 2.7401094008395877, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.451361656188965, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8649752736091614, "num_tokens": 821866160.0, "step": 21540 }, { "epoch": 2.7402366111181786, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.374238014221191, "learning_rate": 1e-06, "loss": 0.3806, "mean_token_accuracy": 0.8804060816764832, "num_tokens": 821905595.0, "step": 21541 }, { "epoch": 2.7403638213967687, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.38539981842041, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8651403188705444, "num_tokens": 821944769.0, "step": 21542 }, { "epoch": 2.7404910316753592, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.381775856018066, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8819862604141235, "num_tokens": 821988889.0, "step": 21543 }, { "epoch": 2.7406182419539498, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.458948135375977, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8773599863052368, "num_tokens": 822025343.0, "step": 21544 }, { "epoch": 2.7407454522325403, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.367226600646973, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8630965948104858, "num_tokens": 822067651.0, "step": 21545 }, { "epoch": 2.740872662511131, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.400676727294922, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8648163676261902, "num_tokens": 822102885.0, "step": 21546 }, { "epoch": 2.7409998727897213, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.364300727844238, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8666276335716248, "num_tokens": 822142515.0, "step": 21547 }, { "epoch": 2.741127083068312, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.39063549041748, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.871474027633667, "num_tokens": 822180745.0, "step": 21548 }, { "epoch": 2.7412542933469024, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.421392440795898, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8819088339805603, "num_tokens": 822215117.0, "step": 21549 }, { "epoch": 2.741381503625493, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.412671089172363, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8647502660751343, "num_tokens": 822258883.0, "step": 21550 }, { "epoch": 2.7415087139040835, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.406933784484863, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8730520009994507, "num_tokens": 822295254.0, "step": 21551 }, { "epoch": 2.741635924182674, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.447591781616211, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8762612342834473, "num_tokens": 822335769.0, "step": 21552 }, { "epoch": 2.7417631344612645, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.364781379699707, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8839931488037109, "num_tokens": 822376873.0, "step": 21553 }, { "epoch": 2.741890344739855, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.428129196166992, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8720232248306274, "num_tokens": 822419868.0, "step": 21554 }, { "epoch": 2.7420175550184456, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.35704517364502, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8731145858764648, "num_tokens": 822456757.0, "step": 21555 }, { "epoch": 2.742144765297036, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.448450088500977, "learning_rate": 1e-06, "loss": 0.3502, "mean_token_accuracy": 0.8873653411865234, "num_tokens": 822492772.0, "step": 21556 }, { "epoch": 2.7422719755756266, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.373393058776855, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8583779335021973, "num_tokens": 822535400.0, "step": 21557 }, { "epoch": 2.742399185854217, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.449539184570312, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8793734312057495, "num_tokens": 822568313.0, "step": 21558 }, { "epoch": 2.7425263961328077, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.378154754638672, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8810898065567017, "num_tokens": 822608487.0, "step": 21559 }, { "epoch": 2.742653606411398, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.329026222229004, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8751762509346008, "num_tokens": 822647938.0, "step": 21560 }, { "epoch": 2.7427808166899883, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.396930694580078, "learning_rate": 1e-06, "loss": 0.3995, "mean_token_accuracy": 0.8701127767562866, "num_tokens": 822685974.0, "step": 21561 }, { "epoch": 2.7429080269685793, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.389972686767578, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8762497305870056, "num_tokens": 822719166.0, "step": 21562 }, { "epoch": 2.7430352372471694, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.382622718811035, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8646202087402344, "num_tokens": 822758364.0, "step": 21563 }, { "epoch": 2.7431624475257603, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.433639526367188, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8823953866958618, "num_tokens": 822798002.0, "step": 21564 }, { "epoch": 2.7432896578043504, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.396805763244629, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.875816285610199, "num_tokens": 822837958.0, "step": 21565 }, { "epoch": 2.743416868082941, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.363137245178223, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8743143081665039, "num_tokens": 822873076.0, "step": 21566 }, { "epoch": 2.7435440783615315, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.389676094055176, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8736119866371155, "num_tokens": 822904693.0, "step": 21567 }, { "epoch": 2.743671288640122, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.377577781677246, "learning_rate": 1e-06, "loss": 0.4072, "mean_token_accuracy": 0.8696931600570679, "num_tokens": 822941192.0, "step": 21568 }, { "epoch": 2.7437984989187125, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.4202299118042, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8623839616775513, "num_tokens": 822982166.0, "step": 21569 }, { "epoch": 2.743925709197303, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.411789894104004, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8685318827629089, "num_tokens": 823021630.0, "step": 21570 }, { "epoch": 2.7440529194758936, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.403519630432129, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8741007447242737, "num_tokens": 823062133.0, "step": 21571 }, { "epoch": 2.744180129754484, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.419910430908203, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8719131350517273, "num_tokens": 823105163.0, "step": 21572 }, { "epoch": 2.7443073400330746, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.35413932800293, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8543453216552734, "num_tokens": 823146878.0, "step": 21573 }, { "epoch": 2.744434550311665, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.447158813476562, "learning_rate": 1e-06, "loss": 0.3452, "mean_token_accuracy": 0.8911572694778442, "num_tokens": 823182216.0, "step": 21574 }, { "epoch": 2.7445617605902557, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.333780288696289, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8823712468147278, "num_tokens": 823227838.0, "step": 21575 }, { "epoch": 2.7446889708688462, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.427833557128906, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8665546178817749, "num_tokens": 823268609.0, "step": 21576 }, { "epoch": 2.7448161811474368, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.385143280029297, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8834489583969116, "num_tokens": 823309426.0, "step": 21577 }, { "epoch": 2.7449433914260273, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.42796802520752, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8693888187408447, "num_tokens": 823346049.0, "step": 21578 }, { "epoch": 2.745070601704618, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.43325424194336, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8684201240539551, "num_tokens": 823378915.0, "step": 21579 }, { "epoch": 2.7451978119832083, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.474156379699707, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8642241954803467, "num_tokens": 823422985.0, "step": 21580 }, { "epoch": 2.745325022261799, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.357996940612793, "learning_rate": 1e-06, "loss": 0.4433, "mean_token_accuracy": 0.8615875840187073, "num_tokens": 823460703.0, "step": 21581 }, { "epoch": 2.7454522325403894, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.444405555725098, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8774727582931519, "num_tokens": 823499651.0, "step": 21582 }, { "epoch": 2.74557944281898, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.392633438110352, "learning_rate": 1e-06, "loss": 0.4059, "mean_token_accuracy": 0.8735747337341309, "num_tokens": 823537044.0, "step": 21583 }, { "epoch": 2.7457066530975704, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.370566368103027, "learning_rate": 1e-06, "loss": 0.3522, "mean_token_accuracy": 0.8866832852363586, "num_tokens": 823571832.0, "step": 21584 }, { "epoch": 2.745833863376161, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.426856994628906, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.8804808855056763, "num_tokens": 823609852.0, "step": 21585 }, { "epoch": 2.745961073654751, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.436930656433105, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8781348466873169, "num_tokens": 823644419.0, "step": 21586 }, { "epoch": 2.746088283933342, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.459002494812012, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8797906637191772, "num_tokens": 823682451.0, "step": 21587 }, { "epoch": 2.746215494211932, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.402166366577148, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.884530246257782, "num_tokens": 823723051.0, "step": 21588 }, { "epoch": 2.746342704490523, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.434972763061523, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8712238669395447, "num_tokens": 823761353.0, "step": 21589 }, { "epoch": 2.746469914769113, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.397974967956543, "learning_rate": 1e-06, "loss": 0.459, "mean_token_accuracy": 0.8605749011039734, "num_tokens": 823800281.0, "step": 21590 }, { "epoch": 2.7465971250477037, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.396500587463379, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8872663974761963, "num_tokens": 823843255.0, "step": 21591 }, { "epoch": 2.7467243353262942, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.442554473876953, "learning_rate": 1e-06, "loss": 0.4045, "mean_token_accuracy": 0.8709242343902588, "num_tokens": 823880736.0, "step": 21592 }, { "epoch": 2.7468515456048848, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42873477935791, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.8721049427986145, "num_tokens": 823923232.0, "step": 21593 }, { "epoch": 2.7469787558834753, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.454864501953125, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8774241209030151, "num_tokens": 823957488.0, "step": 21594 }, { "epoch": 2.747105966162066, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.395635604858398, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8692800402641296, "num_tokens": 823999568.0, "step": 21595 }, { "epoch": 2.7472331764406563, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.517459869384766, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.885852038860321, "num_tokens": 824036947.0, "step": 21596 }, { "epoch": 2.747360386719247, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.462142944335938, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8862379789352417, "num_tokens": 824067966.0, "step": 21597 }, { "epoch": 2.7474875969978374, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.44498062133789, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8735365271568298, "num_tokens": 824112216.0, "step": 21598 }, { "epoch": 2.747614807276428, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.426003456115723, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8847094774246216, "num_tokens": 824148749.0, "step": 21599 }, { "epoch": 2.7477420175550185, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.486746788024902, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8734843730926514, "num_tokens": 824186685.0, "step": 21600 }, { "epoch": 2.747869227833609, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.404601097106934, "learning_rate": 1e-06, "loss": 0.4151, "mean_token_accuracy": 0.8699883222579956, "num_tokens": 824223716.0, "step": 21601 }, { "epoch": 2.7479964381121995, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.507614135742188, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8729602694511414, "num_tokens": 824264627.0, "step": 21602 }, { "epoch": 2.74812364839079, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.435802459716797, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8745995759963989, "num_tokens": 824301901.0, "step": 21603 }, { "epoch": 2.7482508586693806, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.437219619750977, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.877439022064209, "num_tokens": 824344621.0, "step": 21604 }, { "epoch": 2.748378068947971, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.439626693725586, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8793971538543701, "num_tokens": 824385850.0, "step": 21605 }, { "epoch": 2.7485052792265616, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.401123046875, "learning_rate": 1e-06, "loss": 0.3137, "mean_token_accuracy": 0.9015032052993774, "num_tokens": 824425545.0, "step": 21606 }, { "epoch": 2.748632489505152, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.387454986572266, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.8911919593811035, "num_tokens": 824468611.0, "step": 21607 }, { "epoch": 2.7487596997837427, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.448399543762207, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8802517652511597, "num_tokens": 824512161.0, "step": 21608 }, { "epoch": 2.7488869100623328, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.38876724243164, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.887088418006897, "num_tokens": 824555502.0, "step": 21609 }, { "epoch": 2.7490141203409237, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.402605056762695, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.86930251121521, "num_tokens": 824596277.0, "step": 21610 }, { "epoch": 2.749141330619514, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.423462867736816, "learning_rate": 1e-06, "loss": 0.4354, "mean_token_accuracy": 0.8628427982330322, "num_tokens": 824633704.0, "step": 21611 }, { "epoch": 2.749268540898105, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.413434028625488, "learning_rate": 1e-06, "loss": 0.3842, "mean_token_accuracy": 0.8783942461013794, "num_tokens": 824673584.0, "step": 21612 }, { "epoch": 2.749395751176695, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.418651580810547, "learning_rate": 1e-06, "loss": 0.4249, "mean_token_accuracy": 0.868950605392456, "num_tokens": 824712508.0, "step": 21613 }, { "epoch": 2.749522961455286, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.454023361206055, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.8863930702209473, "num_tokens": 824754353.0, "step": 21614 }, { "epoch": 2.749650171733876, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.373080253601074, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8557907342910767, "num_tokens": 824796333.0, "step": 21615 }, { "epoch": 2.7497773820124665, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.449658393859863, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8679870367050171, "num_tokens": 824832048.0, "step": 21616 }, { "epoch": 2.749904592291057, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.476304054260254, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8697862029075623, "num_tokens": 824871889.0, "step": 21617 }, { "epoch": 2.7500318025696475, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.410351753234863, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8843642473220825, "num_tokens": 824912547.0, "step": 21618 }, { "epoch": 2.750159012848238, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.48141860961914, "learning_rate": 1e-06, "loss": 0.4628, "mean_token_accuracy": 0.8557880520820618, "num_tokens": 824955626.0, "step": 21619 }, { "epoch": 2.7502862231268286, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.407803535461426, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.882961630821228, "num_tokens": 824991813.0, "step": 21620 }, { "epoch": 2.750413433405419, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.444812774658203, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8826727867126465, "num_tokens": 825024356.0, "step": 21621 }, { "epoch": 2.7505406436840096, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.38862133026123, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8719255328178406, "num_tokens": 825060316.0, "step": 21622 }, { "epoch": 2.7506678539626, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.419279098510742, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8678241968154907, "num_tokens": 825096044.0, "step": 21623 }, { "epoch": 2.7507950642411907, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45312213897705, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8722942471504211, "num_tokens": 825123999.0, "step": 21624 }, { "epoch": 2.750922274519781, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.397021293640137, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8642787933349609, "num_tokens": 825170155.0, "step": 21625 }, { "epoch": 2.7510494847983717, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.374517440795898, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8738190531730652, "num_tokens": 825213388.0, "step": 21626 }, { "epoch": 2.7511766950769623, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.407989501953125, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8811593651771545, "num_tokens": 825247171.0, "step": 21627 }, { "epoch": 2.751303905355553, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.486655235290527, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8758933544158936, "num_tokens": 825283271.0, "step": 21628 }, { "epoch": 2.7514311156341433, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.40860652923584, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8735995292663574, "num_tokens": 825318976.0, "step": 21629 }, { "epoch": 2.751558325912734, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.404667854309082, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8738186359405518, "num_tokens": 825358019.0, "step": 21630 }, { "epoch": 2.7516855361913244, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.458991050720215, "learning_rate": 1e-06, "loss": 0.3564, "mean_token_accuracy": 0.8911879062652588, "num_tokens": 825394025.0, "step": 21631 }, { "epoch": 2.751812746469915, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.446971893310547, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8855357766151428, "num_tokens": 825433021.0, "step": 21632 }, { "epoch": 2.7519399567485054, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.39830207824707, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8753143548965454, "num_tokens": 825477026.0, "step": 21633 }, { "epoch": 2.7520671670270955, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.400688171386719, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8787543773651123, "num_tokens": 825511054.0, "step": 21634 }, { "epoch": 2.7521943773056865, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.398530006408691, "learning_rate": 1e-06, "loss": 0.4368, "mean_token_accuracy": 0.8607957363128662, "num_tokens": 825549654.0, "step": 21635 }, { "epoch": 2.7523215875842766, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.408125877380371, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8771061897277832, "num_tokens": 825587269.0, "step": 21636 }, { "epoch": 2.7524487978628676, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.395635604858398, "learning_rate": 1e-06, "loss": 0.3264, "mean_token_accuracy": 0.8975348472595215, "num_tokens": 825624864.0, "step": 21637 }, { "epoch": 2.7525760081414576, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.398307800292969, "learning_rate": 1e-06, "loss": 0.4294, "mean_token_accuracy": 0.8677629232406616, "num_tokens": 825667025.0, "step": 21638 }, { "epoch": 2.7527032184200486, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.419639587402344, "learning_rate": 1e-06, "loss": 0.4425, "mean_token_accuracy": 0.8595465421676636, "num_tokens": 825709505.0, "step": 21639 }, { "epoch": 2.7528304286986387, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.423377990722656, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8779785633087158, "num_tokens": 825747376.0, "step": 21640 }, { "epoch": 2.7529576389772292, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.3941011428833, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8810791969299316, "num_tokens": 825784817.0, "step": 21641 }, { "epoch": 2.7530848492558198, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.441814422607422, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8731930255889893, "num_tokens": 825815631.0, "step": 21642 }, { "epoch": 2.7532120595344103, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.408778190612793, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8749412298202515, "num_tokens": 825854202.0, "step": 21643 }, { "epoch": 2.753339269813001, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.439441680908203, "learning_rate": 1e-06, "loss": 0.3761, "mean_token_accuracy": 0.8831206560134888, "num_tokens": 825890636.0, "step": 21644 }, { "epoch": 2.7534664800915913, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.444446563720703, "learning_rate": 1e-06, "loss": 0.4572, "mean_token_accuracy": 0.853023886680603, "num_tokens": 825931054.0, "step": 21645 }, { "epoch": 2.753593690370182, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.398773193359375, "learning_rate": 1e-06, "loss": 0.409, "mean_token_accuracy": 0.8707168102264404, "num_tokens": 825972519.0, "step": 21646 }, { "epoch": 2.7537209006487724, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.404105186462402, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8766008615493774, "num_tokens": 826010010.0, "step": 21647 }, { "epoch": 2.753848110927363, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42053508758545, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8743489980697632, "num_tokens": 826048348.0, "step": 21648 }, { "epoch": 2.7539753212059535, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.438831329345703, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.870895266532898, "num_tokens": 826087807.0, "step": 21649 }, { "epoch": 2.754102531484544, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.432456016540527, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8707945346832275, "num_tokens": 826124657.0, "step": 21650 }, { "epoch": 2.7542297417631345, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.347105979919434, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8764652609825134, "num_tokens": 826164190.0, "step": 21651 }, { "epoch": 2.754356952041725, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.403428077697754, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8761382102966309, "num_tokens": 826202876.0, "step": 21652 }, { "epoch": 2.7544841623203156, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.459756851196289, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8737056255340576, "num_tokens": 826242955.0, "step": 21653 }, { "epoch": 2.754611372598906, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.416842460632324, "learning_rate": 1e-06, "loss": 0.3665, "mean_token_accuracy": 0.8830679059028625, "num_tokens": 826279925.0, "step": 21654 }, { "epoch": 2.7547385828774966, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.379611015319824, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8847737312316895, "num_tokens": 826314076.0, "step": 21655 }, { "epoch": 2.754865793156087, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.434900283813477, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8720346093177795, "num_tokens": 826354138.0, "step": 21656 }, { "epoch": 2.7549930034346777, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.346787452697754, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8708702325820923, "num_tokens": 826395393.0, "step": 21657 }, { "epoch": 2.755120213713268, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.43961238861084, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8653780817985535, "num_tokens": 826437539.0, "step": 21658 }, { "epoch": 2.7552474239918583, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.388798713684082, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8680963516235352, "num_tokens": 826475857.0, "step": 21659 }, { "epoch": 2.7553746342704493, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.389808654785156, "learning_rate": 1e-06, "loss": 0.3747, "mean_token_accuracy": 0.8834657073020935, "num_tokens": 826513956.0, "step": 21660 }, { "epoch": 2.7555018445490393, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.409285545349121, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8732596635818481, "num_tokens": 826552747.0, "step": 21661 }, { "epoch": 2.7556290548276303, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.369806289672852, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8815013766288757, "num_tokens": 826590830.0, "step": 21662 }, { "epoch": 2.7557562651062204, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.370471954345703, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8773195743560791, "num_tokens": 826633414.0, "step": 21663 }, { "epoch": 2.755883475384811, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.396775245666504, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8835672736167908, "num_tokens": 826668344.0, "step": 21664 }, { "epoch": 2.7560106856634015, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.424066543579102, "learning_rate": 1e-06, "loss": 0.4531, "mean_token_accuracy": 0.8575493693351746, "num_tokens": 826710587.0, "step": 21665 }, { "epoch": 2.756137895941992, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.401538848876953, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8648152351379395, "num_tokens": 826748854.0, "step": 21666 }, { "epoch": 2.7562651062205825, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.403820991516113, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8616167902946472, "num_tokens": 826789591.0, "step": 21667 }, { "epoch": 2.756392316499173, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.370488166809082, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8709500432014465, "num_tokens": 826831890.0, "step": 21668 }, { "epoch": 2.7565195267777636, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.403573989868164, "learning_rate": 1e-06, "loss": 0.4533, "mean_token_accuracy": 0.8576011657714844, "num_tokens": 826870265.0, "step": 21669 }, { "epoch": 2.756646737056354, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.425532341003418, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8737233877182007, "num_tokens": 826910993.0, "step": 21670 }, { "epoch": 2.7567739473349446, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.394364356994629, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8728109002113342, "num_tokens": 826947492.0, "step": 21671 }, { "epoch": 2.756901157613535, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.464920043945312, "learning_rate": 1e-06, "loss": 0.3889, "mean_token_accuracy": 0.8763192296028137, "num_tokens": 826987215.0, "step": 21672 }, { "epoch": 2.7570283678921257, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.426679611206055, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8721016645431519, "num_tokens": 827024092.0, "step": 21673 }, { "epoch": 2.757155578170716, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.41375732421875, "learning_rate": 1e-06, "loss": 0.4614, "mean_token_accuracy": 0.858324408531189, "num_tokens": 827060541.0, "step": 21674 }, { "epoch": 2.7572827884493067, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.440062522888184, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8731070756912231, "num_tokens": 827101205.0, "step": 21675 }, { "epoch": 2.7574099987278973, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.43968677520752, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8681012988090515, "num_tokens": 827132441.0, "step": 21676 }, { "epoch": 2.757537209006488, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42782211303711, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8747077584266663, "num_tokens": 827169530.0, "step": 21677 }, { "epoch": 2.7576644192850783, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.364839553833008, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8805224299430847, "num_tokens": 827207352.0, "step": 21678 }, { "epoch": 2.757791629563669, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.438048362731934, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8686591386795044, "num_tokens": 827242790.0, "step": 21679 }, { "epoch": 2.7579188398422594, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.356358528137207, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8752266764640808, "num_tokens": 827281676.0, "step": 21680 }, { "epoch": 2.75804605012085, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.442651748657227, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8878270983695984, "num_tokens": 827313676.0, "step": 21681 }, { "epoch": 2.7581732603994404, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.361997604370117, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8744628429412842, "num_tokens": 827355017.0, "step": 21682 }, { "epoch": 2.758300470678031, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.431127548217773, "learning_rate": 1e-06, "loss": 0.3501, "mean_token_accuracy": 0.8914260864257812, "num_tokens": 827391773.0, "step": 21683 }, { "epoch": 2.758427680956621, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.391923904418945, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8889837861061096, "num_tokens": 827433034.0, "step": 21684 }, { "epoch": 2.758554891235212, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.376047134399414, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8835965991020203, "num_tokens": 827475729.0, "step": 21685 }, { "epoch": 2.758682101513802, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.372587203979492, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8793950080871582, "num_tokens": 827514127.0, "step": 21686 }, { "epoch": 2.758809311792393, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.425321578979492, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8847751617431641, "num_tokens": 827546709.0, "step": 21687 }, { "epoch": 2.758936522070983, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.457112312316895, "learning_rate": 1e-06, "loss": 0.3343, "mean_token_accuracy": 0.8945344686508179, "num_tokens": 827581450.0, "step": 21688 }, { "epoch": 2.7590637323495737, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.432217597961426, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8706650137901306, "num_tokens": 827623161.0, "step": 21689 }, { "epoch": 2.7591909426281642, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.376468658447266, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8704062700271606, "num_tokens": 827659460.0, "step": 21690 }, { "epoch": 2.7593181529067548, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.450228691101074, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8695712685585022, "num_tokens": 827701110.0, "step": 21691 }, { "epoch": 2.7594453631853453, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.45533275604248, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8714594841003418, "num_tokens": 827738992.0, "step": 21692 }, { "epoch": 2.759572573463936, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.452386856079102, "learning_rate": 1e-06, "loss": 0.4417, "mean_token_accuracy": 0.8602758049964905, "num_tokens": 827775853.0, "step": 21693 }, { "epoch": 2.7596997837425263, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.399975776672363, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8762781620025635, "num_tokens": 827819438.0, "step": 21694 }, { "epoch": 2.759826994021117, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.396208763122559, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8689752817153931, "num_tokens": 827858833.0, "step": 21695 }, { "epoch": 2.7599542042997074, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.454842567443848, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8676941394805908, "num_tokens": 827895069.0, "step": 21696 }, { "epoch": 2.760081414578298, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.398975372314453, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8878474235534668, "num_tokens": 827932970.0, "step": 21697 }, { "epoch": 2.7602086248568884, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.410124778747559, "learning_rate": 1e-06, "loss": 0.3639, "mean_token_accuracy": 0.886118471622467, "num_tokens": 827981859.0, "step": 21698 }, { "epoch": 2.760335835135479, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.4230318069458, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8684666752815247, "num_tokens": 828022018.0, "step": 21699 }, { "epoch": 2.7604630454140695, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.456138610839844, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8702597618103027, "num_tokens": 828061546.0, "step": 21700 }, { "epoch": 2.76059025569266, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.414471626281738, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8611669540405273, "num_tokens": 828104464.0, "step": 21701 }, { "epoch": 2.7607174659712506, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.434934616088867, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8638144135475159, "num_tokens": 828142527.0, "step": 21702 }, { "epoch": 2.760844676249841, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.41973876953125, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8748518228530884, "num_tokens": 828189670.0, "step": 21703 }, { "epoch": 2.7609718865284316, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.410155296325684, "learning_rate": 1e-06, "loss": 0.379, "mean_token_accuracy": 0.8809897303581238, "num_tokens": 828231511.0, "step": 21704 }, { "epoch": 2.761099096807022, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.371102333068848, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.8742812871932983, "num_tokens": 828274387.0, "step": 21705 }, { "epoch": 2.7612263070856127, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.424962997436523, "learning_rate": 1e-06, "loss": 0.4163, "mean_token_accuracy": 0.8670259118080139, "num_tokens": 828312263.0, "step": 21706 }, { "epoch": 2.7613535173642028, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.382984161376953, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.875694990158081, "num_tokens": 828345550.0, "step": 21707 }, { "epoch": 2.7614807276427937, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.456963539123535, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8639583587646484, "num_tokens": 828378633.0, "step": 21708 }, { "epoch": 2.761607937921384, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.43358039855957, "learning_rate": 1e-06, "loss": 0.421, "mean_token_accuracy": 0.8681100606918335, "num_tokens": 828414292.0, "step": 21709 }, { "epoch": 2.761735148199975, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.394527435302734, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8730829954147339, "num_tokens": 828452682.0, "step": 21710 }, { "epoch": 2.761862358478565, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.433213233947754, "learning_rate": 1e-06, "loss": 0.3554, "mean_token_accuracy": 0.8897738456726074, "num_tokens": 828494162.0, "step": 21711 }, { "epoch": 2.761989568757156, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.470192909240723, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8787976503372192, "num_tokens": 828525997.0, "step": 21712 }, { "epoch": 2.762116779035746, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.460787773132324, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8728806376457214, "num_tokens": 828562488.0, "step": 21713 }, { "epoch": 2.7622439893143365, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.403106689453125, "learning_rate": 1e-06, "loss": 0.3432, "mean_token_accuracy": 0.8922480940818787, "num_tokens": 828591818.0, "step": 21714 }, { "epoch": 2.762371199592927, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.436405181884766, "learning_rate": 1e-06, "loss": 0.4379, "mean_token_accuracy": 0.8631081581115723, "num_tokens": 828628872.0, "step": 21715 }, { "epoch": 2.7624984098715175, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.422176361083984, "learning_rate": 1e-06, "loss": 0.3772, "mean_token_accuracy": 0.8829811811447144, "num_tokens": 828670939.0, "step": 21716 }, { "epoch": 2.762625620150108, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.463876724243164, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8569382429122925, "num_tokens": 828708446.0, "step": 21717 }, { "epoch": 2.7627528304286986, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.363656997680664, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8784293532371521, "num_tokens": 828749873.0, "step": 21718 }, { "epoch": 2.762880040707289, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.401721954345703, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.8679863810539246, "num_tokens": 828790066.0, "step": 21719 }, { "epoch": 2.7630072509858796, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.40148639678955, "learning_rate": 1e-06, "loss": 0.4202, "mean_token_accuracy": 0.8708337545394897, "num_tokens": 828821911.0, "step": 21720 }, { "epoch": 2.76313446126447, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.48420238494873, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8690462112426758, "num_tokens": 828856572.0, "step": 21721 }, { "epoch": 2.7632616715430607, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.392560005187988, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8846603631973267, "num_tokens": 828898865.0, "step": 21722 }, { "epoch": 2.763388881821651, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.40038013458252, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8818466067314148, "num_tokens": 828933768.0, "step": 21723 }, { "epoch": 2.7635160921002417, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.455462455749512, "learning_rate": 1e-06, "loss": 0.3544, "mean_token_accuracy": 0.8893228769302368, "num_tokens": 828968869.0, "step": 21724 }, { "epoch": 2.7636433023788323, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.439229965209961, "learning_rate": 1e-06, "loss": 0.4602, "mean_token_accuracy": 0.8587903380393982, "num_tokens": 829005621.0, "step": 21725 }, { "epoch": 2.763770512657423, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.40660572052002, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8641883134841919, "num_tokens": 829045046.0, "step": 21726 }, { "epoch": 2.7638977229360133, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.40333080291748, "learning_rate": 1e-06, "loss": 0.371, "mean_token_accuracy": 0.8872920274734497, "num_tokens": 829084588.0, "step": 21727 }, { "epoch": 2.764024933214604, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.457204818725586, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8583807945251465, "num_tokens": 829121010.0, "step": 21728 }, { "epoch": 2.7641521434931944, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.414292335510254, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8827428221702576, "num_tokens": 829159618.0, "step": 21729 }, { "epoch": 2.764279353771785, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.453052520751953, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8617969751358032, "num_tokens": 829204826.0, "step": 21730 }, { "epoch": 2.7644065640503754, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4285249710083, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.8846818208694458, "num_tokens": 829240741.0, "step": 21731 }, { "epoch": 2.7645337743289655, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45255184173584, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.8926066160202026, "num_tokens": 829278065.0, "step": 21732 }, { "epoch": 2.7646609846075565, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.467167854309082, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8699589967727661, "num_tokens": 829319438.0, "step": 21733 }, { "epoch": 2.7647881948861466, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.40671157836914, "learning_rate": 1e-06, "loss": 0.3702, "mean_token_accuracy": 0.8844167590141296, "num_tokens": 829357299.0, "step": 21734 }, { "epoch": 2.7649154051647375, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.508378028869629, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8752198219299316, "num_tokens": 829398600.0, "step": 21735 }, { "epoch": 2.7650426154433276, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.411208152770996, "learning_rate": 1e-06, "loss": 0.4311, "mean_token_accuracy": 0.8630760908126831, "num_tokens": 829439214.0, "step": 21736 }, { "epoch": 2.7651698257219186, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.468515396118164, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8638729453086853, "num_tokens": 829477634.0, "step": 21737 }, { "epoch": 2.7652970360005087, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.389703750610352, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8811020255088806, "num_tokens": 829514527.0, "step": 21738 }, { "epoch": 2.765424246279099, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.44884204864502, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.87591552734375, "num_tokens": 829558531.0, "step": 21739 }, { "epoch": 2.7655514565576897, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.414179801940918, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.8770676851272583, "num_tokens": 829600769.0, "step": 21740 }, { "epoch": 2.7656786668362803, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.433395385742188, "learning_rate": 1e-06, "loss": 0.3854, "mean_token_accuracy": 0.8783622980117798, "num_tokens": 829635938.0, "step": 21741 }, { "epoch": 2.765805877114871, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.464573860168457, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8749486207962036, "num_tokens": 829672956.0, "step": 21742 }, { "epoch": 2.7659330873934613, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.449407577514648, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8739639520645142, "num_tokens": 829712658.0, "step": 21743 }, { "epoch": 2.766060297672052, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.424463272094727, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8775732517242432, "num_tokens": 829755429.0, "step": 21744 }, { "epoch": 2.7661875079506424, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.412473678588867, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8737466335296631, "num_tokens": 829796373.0, "step": 21745 }, { "epoch": 2.766314718229233, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.452178955078125, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8697645664215088, "num_tokens": 829834049.0, "step": 21746 }, { "epoch": 2.7664419285078234, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.429608345031738, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8744092583656311, "num_tokens": 829870853.0, "step": 21747 }, { "epoch": 2.766569138786414, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.48979663848877, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8679137229919434, "num_tokens": 829907519.0, "step": 21748 }, { "epoch": 2.7666963490650045, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.393155097961426, "learning_rate": 1e-06, "loss": 0.4348, "mean_token_accuracy": 0.8611283302307129, "num_tokens": 829947162.0, "step": 21749 }, { "epoch": 2.766823559343595, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.459357261657715, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8726402521133423, "num_tokens": 829991981.0, "step": 21750 }, { "epoch": 2.7669507696221856, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.407414436340332, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8799384236335754, "num_tokens": 830032409.0, "step": 21751 }, { "epoch": 2.767077979900776, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.386835098266602, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8775228261947632, "num_tokens": 830072738.0, "step": 21752 }, { "epoch": 2.7672051901793666, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.399503707885742, "learning_rate": 1e-06, "loss": 0.4128, "mean_token_accuracy": 0.8709055185317993, "num_tokens": 830110880.0, "step": 21753 }, { "epoch": 2.767332400457957, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.43130874633789, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.8693050146102905, "num_tokens": 830150148.0, "step": 21754 }, { "epoch": 2.7674596107365477, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.406228065490723, "learning_rate": 1e-06, "loss": 0.4445, "mean_token_accuracy": 0.8627656102180481, "num_tokens": 830190717.0, "step": 21755 }, { "epoch": 2.767586821015138, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.481679916381836, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8755241632461548, "num_tokens": 830226972.0, "step": 21756 }, { "epoch": 2.7677140312937283, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.499367713928223, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8827057480812073, "num_tokens": 830262612.0, "step": 21757 }, { "epoch": 2.7678412415723193, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.352165222167969, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.880463719367981, "num_tokens": 830306856.0, "step": 21758 }, { "epoch": 2.7679684518509093, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.46352481842041, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8724529147148132, "num_tokens": 830340726.0, "step": 21759 }, { "epoch": 2.7680956621295003, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.390697479248047, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.873505711555481, "num_tokens": 830380697.0, "step": 21760 }, { "epoch": 2.7682228724080904, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.535470962524414, "learning_rate": 1e-06, "loss": 0.4512, "mean_token_accuracy": 0.8526347875595093, "num_tokens": 830414737.0, "step": 21761 }, { "epoch": 2.768350082686681, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.429986000061035, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8742179870605469, "num_tokens": 830456518.0, "step": 21762 }, { "epoch": 2.7684772929652715, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45946979522705, "learning_rate": 1e-06, "loss": 0.407, "mean_token_accuracy": 0.8707338571548462, "num_tokens": 830494192.0, "step": 21763 }, { "epoch": 2.768604503243862, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.454398155212402, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8673735857009888, "num_tokens": 830533659.0, "step": 21764 }, { "epoch": 2.7687317135224525, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.447633743286133, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8682454228401184, "num_tokens": 830569605.0, "step": 21765 }, { "epoch": 2.768858923801043, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.492514610290527, "learning_rate": 1e-06, "loss": 0.4648, "mean_token_accuracy": 0.8526269197463989, "num_tokens": 830610814.0, "step": 21766 }, { "epoch": 2.7689861340796336, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.426190376281738, "learning_rate": 1e-06, "loss": 0.4053, "mean_token_accuracy": 0.8729468584060669, "num_tokens": 830646811.0, "step": 21767 }, { "epoch": 2.769113344358224, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.458056449890137, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.877022385597229, "num_tokens": 830680984.0, "step": 21768 }, { "epoch": 2.7692405546368146, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.481627464294434, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.873663067817688, "num_tokens": 830713384.0, "step": 21769 }, { "epoch": 2.769367764915405, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.436773300170898, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8611578941345215, "num_tokens": 830746922.0, "step": 21770 }, { "epoch": 2.7694949751939957, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.454282760620117, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.8797656297683716, "num_tokens": 830789459.0, "step": 21771 }, { "epoch": 2.769622185472586, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.450407028198242, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8709287047386169, "num_tokens": 830826736.0, "step": 21772 }, { "epoch": 2.7697493957511767, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.442310333251953, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8764512538909912, "num_tokens": 830868118.0, "step": 21773 }, { "epoch": 2.7698766060297673, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.466673851013184, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8702648878097534, "num_tokens": 830911658.0, "step": 21774 }, { "epoch": 2.770003816308358, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.431965827941895, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8749613761901855, "num_tokens": 830948456.0, "step": 21775 }, { "epoch": 2.7701310265869483, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.446261405944824, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8832637071609497, "num_tokens": 830988790.0, "step": 21776 }, { "epoch": 2.770258236865539, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.4219970703125, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8665924072265625, "num_tokens": 831024551.0, "step": 21777 }, { "epoch": 2.7703854471441294, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.488191604614258, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8711954951286316, "num_tokens": 831060082.0, "step": 21778 }, { "epoch": 2.77051265742272, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.432952880859375, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.874548077583313, "num_tokens": 831100067.0, "step": 21779 }, { "epoch": 2.77063986770131, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.483540534973145, "learning_rate": 1e-06, "loss": 0.3638, "mean_token_accuracy": 0.8850656747817993, "num_tokens": 831131923.0, "step": 21780 }, { "epoch": 2.770767077979901, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.46618938446045, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8591698408126831, "num_tokens": 831167985.0, "step": 21781 }, { "epoch": 2.770894288258491, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.44529914855957, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8690038323402405, "num_tokens": 831211737.0, "step": 21782 }, { "epoch": 2.771021498537082, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.461000442504883, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8635708689689636, "num_tokens": 831248764.0, "step": 21783 }, { "epoch": 2.771148708815672, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.455557823181152, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.8799923658370972, "num_tokens": 831284140.0, "step": 21784 }, { "epoch": 2.771275919094263, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.466286659240723, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8764945864677429, "num_tokens": 831318567.0, "step": 21785 }, { "epoch": 2.771403129372853, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.430777549743652, "learning_rate": 1e-06, "loss": 0.4179, "mean_token_accuracy": 0.8675814867019653, "num_tokens": 831357128.0, "step": 21786 }, { "epoch": 2.7715303396514437, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4977388381958, "learning_rate": 1e-06, "loss": 0.4701, "mean_token_accuracy": 0.8544707298278809, "num_tokens": 831391551.0, "step": 21787 }, { "epoch": 2.771657549930034, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.44750690460205, "learning_rate": 1e-06, "loss": 0.4172, "mean_token_accuracy": 0.8698956370353699, "num_tokens": 831430332.0, "step": 21788 }, { "epoch": 2.7717847602086247, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4953031539917, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8790240287780762, "num_tokens": 831471639.0, "step": 21789 }, { "epoch": 2.7719119704872153, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.469160079956055, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8786265254020691, "num_tokens": 831506470.0, "step": 21790 }, { "epoch": 2.772039180765806, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.499513626098633, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8668603897094727, "num_tokens": 831547212.0, "step": 21791 }, { "epoch": 2.7721663910443963, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.463907241821289, "learning_rate": 1e-06, "loss": 0.4419, "mean_token_accuracy": 0.8587411046028137, "num_tokens": 831593181.0, "step": 21792 }, { "epoch": 2.772293601322987, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.437796592712402, "learning_rate": 1e-06, "loss": 0.3419, "mean_token_accuracy": 0.8902103304862976, "num_tokens": 831624889.0, "step": 21793 }, { "epoch": 2.7724208116015774, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45101261138916, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8795157670974731, "num_tokens": 831662389.0, "step": 21794 }, { "epoch": 2.772548021880168, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.493894577026367, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8753217458724976, "num_tokens": 831705282.0, "step": 21795 }, { "epoch": 2.7726752321587584, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.468636512756348, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8801196813583374, "num_tokens": 831741604.0, "step": 21796 }, { "epoch": 2.772802442437349, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.523932456970215, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8649823069572449, "num_tokens": 831784165.0, "step": 21797 }, { "epoch": 2.7729296527159395, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.498298645019531, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8785585165023804, "num_tokens": 831821101.0, "step": 21798 }, { "epoch": 2.77305686299453, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.446949005126953, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8765360116958618, "num_tokens": 831855766.0, "step": 21799 }, { "epoch": 2.7731840732731206, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.477584838867188, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8794155120849609, "num_tokens": 831895692.0, "step": 21800 }, { "epoch": 2.773311283551711, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.477959632873535, "learning_rate": 1e-06, "loss": 0.334, "mean_token_accuracy": 0.8925188779830933, "num_tokens": 831931079.0, "step": 21801 }, { "epoch": 2.7734384938303016, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.486929893493652, "learning_rate": 1e-06, "loss": 0.4731, "mean_token_accuracy": 0.8492304086685181, "num_tokens": 831969742.0, "step": 21802 }, { "epoch": 2.773565704108892, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.487014770507812, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8636721968650818, "num_tokens": 832011059.0, "step": 21803 }, { "epoch": 2.7736929143874827, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.493565559387207, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.883423924446106, "num_tokens": 832042768.0, "step": 21804 }, { "epoch": 2.7738201246660728, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.499991416931152, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8826354146003723, "num_tokens": 832077699.0, "step": 21805 }, { "epoch": 2.7739473349446637, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.485697746276855, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8812035918235779, "num_tokens": 832113528.0, "step": 21806 }, { "epoch": 2.774074545223254, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.499424934387207, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8779171705245972, "num_tokens": 832149108.0, "step": 21807 }, { "epoch": 2.774201755501845, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.439461708068848, "learning_rate": 1e-06, "loss": 0.3526, "mean_token_accuracy": 0.8905742764472961, "num_tokens": 832189279.0, "step": 21808 }, { "epoch": 2.774328965780435, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.508305549621582, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.87546306848526, "num_tokens": 832224423.0, "step": 21809 }, { "epoch": 2.774456176059026, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.484089851379395, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8680417537689209, "num_tokens": 832260389.0, "step": 21810 }, { "epoch": 2.774583386337616, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.418723106384277, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8732805252075195, "num_tokens": 832297591.0, "step": 21811 }, { "epoch": 2.7747105966162064, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.44651985168457, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8790022730827332, "num_tokens": 832330195.0, "step": 21812 }, { "epoch": 2.774837806894797, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.525472640991211, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8812936544418335, "num_tokens": 832361613.0, "step": 21813 }, { "epoch": 2.7749650171733875, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.453893661499023, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8798304200172424, "num_tokens": 832402800.0, "step": 21814 }, { "epoch": 2.775092227451978, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.418891906738281, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8764320015907288, "num_tokens": 832445161.0, "step": 21815 }, { "epoch": 2.7752194377305686, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.43917465209961, "learning_rate": 1e-06, "loss": 0.437, "mean_token_accuracy": 0.8600151538848877, "num_tokens": 832487905.0, "step": 21816 }, { "epoch": 2.775346648009159, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45494270324707, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8758556842803955, "num_tokens": 832521107.0, "step": 21817 }, { "epoch": 2.7754738582877496, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.449894905090332, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8703808784484863, "num_tokens": 832557213.0, "step": 21818 }, { "epoch": 2.77560106856634, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.452460289001465, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8619207739830017, "num_tokens": 832600445.0, "step": 21819 }, { "epoch": 2.7757282788449307, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.41866683959961, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8733255863189697, "num_tokens": 832635994.0, "step": 21820 }, { "epoch": 2.775855489123521, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.468778610229492, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8846750259399414, "num_tokens": 832677151.0, "step": 21821 }, { "epoch": 2.7759826994021117, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.438519477844238, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8739737868309021, "num_tokens": 832713995.0, "step": 21822 }, { "epoch": 2.7761099096807023, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.433069229125977, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8652689456939697, "num_tokens": 832754201.0, "step": 21823 }, { "epoch": 2.776237119959293, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.433006286621094, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8707638382911682, "num_tokens": 832796920.0, "step": 21824 }, { "epoch": 2.7763643302378833, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.415082931518555, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8705277442932129, "num_tokens": 832835952.0, "step": 21825 }, { "epoch": 2.776491540516474, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.452866554260254, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8769075870513916, "num_tokens": 832878306.0, "step": 21826 }, { "epoch": 2.7766187507950644, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.434473037719727, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8668990135192871, "num_tokens": 832914315.0, "step": 21827 }, { "epoch": 2.776745961073655, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4158935546875, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8701924085617065, "num_tokens": 832956661.0, "step": 21828 }, { "epoch": 2.7768731713522454, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.49239444732666, "learning_rate": 1e-06, "loss": 0.3805, "mean_token_accuracy": 0.8796898126602173, "num_tokens": 832999446.0, "step": 21829 }, { "epoch": 2.7770003816308355, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.485247611999512, "learning_rate": 1e-06, "loss": 0.4741, "mean_token_accuracy": 0.8467935919761658, "num_tokens": 833039001.0, "step": 21830 }, { "epoch": 2.7771275919094265, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.470253944396973, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8722237944602966, "num_tokens": 833074334.0, "step": 21831 }, { "epoch": 2.7772548021880166, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.494046211242676, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8618516325950623, "num_tokens": 833113089.0, "step": 21832 }, { "epoch": 2.7773820124666075, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42343807220459, "learning_rate": 1e-06, "loss": 0.4261, "mean_token_accuracy": 0.8707174062728882, "num_tokens": 833149755.0, "step": 21833 }, { "epoch": 2.7775092227451976, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.481552124023438, "learning_rate": 1e-06, "loss": 0.4405, "mean_token_accuracy": 0.8647293448448181, "num_tokens": 833187718.0, "step": 21834 }, { "epoch": 2.7776364330237886, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.475085258483887, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8808963894844055, "num_tokens": 833221416.0, "step": 21835 }, { "epoch": 2.7777636433023787, "ewc_loss": 0.039306640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 14.405306816101074, "learning_rate": 1e-06, "loss": 0.3611, "mean_token_accuracy": 0.8873652219772339, "num_tokens": 833261101.0, "step": 21836 }, { "epoch": 2.777890853580969, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.470399856567383, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8822739720344543, "num_tokens": 833296804.0, "step": 21837 }, { "epoch": 2.7780180638595597, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.432360649108887, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.874165415763855, "num_tokens": 833332005.0, "step": 21838 }, { "epoch": 2.7781452741381503, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.453718185424805, "learning_rate": 1e-06, "loss": 0.4112, "mean_token_accuracy": 0.8700626492500305, "num_tokens": 833371991.0, "step": 21839 }, { "epoch": 2.778272484416741, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.459388732910156, "learning_rate": 1e-06, "loss": 0.3651, "mean_token_accuracy": 0.885581374168396, "num_tokens": 833407514.0, "step": 21840 }, { "epoch": 2.7783996946953313, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.406761169433594, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.872665524482727, "num_tokens": 833444519.0, "step": 21841 }, { "epoch": 2.778526904973922, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.501723289489746, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8698252439498901, "num_tokens": 833482252.0, "step": 21842 }, { "epoch": 2.7786541152525124, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.492796897888184, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.872490644454956, "num_tokens": 833515513.0, "step": 21843 }, { "epoch": 2.778781325531103, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.456878662109375, "learning_rate": 1e-06, "loss": 0.3832, "mean_token_accuracy": 0.8804821968078613, "num_tokens": 833551188.0, "step": 21844 }, { "epoch": 2.7789085358096934, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.435776710510254, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8834695816040039, "num_tokens": 833589159.0, "step": 21845 }, { "epoch": 2.779035746088284, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.536652565002441, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8643314242362976, "num_tokens": 833623581.0, "step": 21846 }, { "epoch": 2.7791629563668745, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.444578170776367, "learning_rate": 1e-06, "loss": 0.4358, "mean_token_accuracy": 0.862236499786377, "num_tokens": 833660466.0, "step": 21847 }, { "epoch": 2.779290166645465, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.46413803100586, "learning_rate": 1e-06, "loss": 0.4396, "mean_token_accuracy": 0.8633366227149963, "num_tokens": 833697465.0, "step": 21848 }, { "epoch": 2.7794173769240555, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.43443775177002, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8708842992782593, "num_tokens": 833735178.0, "step": 21849 }, { "epoch": 2.779544587202646, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.423711776733398, "learning_rate": 1e-06, "loss": 0.3488, "mean_token_accuracy": 0.8906490206718445, "num_tokens": 833770193.0, "step": 21850 }, { "epoch": 2.7796717974812366, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.483701705932617, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8741032481193542, "num_tokens": 833807836.0, "step": 21851 }, { "epoch": 2.779799007759827, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.446637153625488, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8712192177772522, "num_tokens": 833847605.0, "step": 21852 }, { "epoch": 2.7799262180384177, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.454530715942383, "learning_rate": 1e-06, "loss": 0.3914, "mean_token_accuracy": 0.8771349787712097, "num_tokens": 833884239.0, "step": 21853 }, { "epoch": 2.780053428317008, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.491056442260742, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.874085009098053, "num_tokens": 833921305.0, "step": 21854 }, { "epoch": 2.7801806385955983, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.467949867248535, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8654682636260986, "num_tokens": 833959297.0, "step": 21855 }, { "epoch": 2.7803078488741892, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.4861421585083, "learning_rate": 1e-06, "loss": 0.4739, "mean_token_accuracy": 0.8519234657287598, "num_tokens": 833992135.0, "step": 21856 }, { "epoch": 2.7804350591527793, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.488800048828125, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8643240928649902, "num_tokens": 834035933.0, "step": 21857 }, { "epoch": 2.7805622694313703, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.447092056274414, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8735895156860352, "num_tokens": 834070187.0, "step": 21858 }, { "epoch": 2.7806894797099604, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417176246643066, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8747307062149048, "num_tokens": 834107370.0, "step": 21859 }, { "epoch": 2.780816689988551, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.425568580627441, "learning_rate": 1e-06, "loss": 0.4191, "mean_token_accuracy": 0.86579430103302, "num_tokens": 834145357.0, "step": 21860 }, { "epoch": 2.7809439002671414, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.432692527770996, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8711419105529785, "num_tokens": 834187250.0, "step": 21861 }, { "epoch": 2.781071110545732, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.477801322937012, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8841845393180847, "num_tokens": 834227598.0, "step": 21862 }, { "epoch": 2.7811983208243225, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417465209960938, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8879445195198059, "num_tokens": 834267032.0, "step": 21863 }, { "epoch": 2.781325531102913, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.439482688903809, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8631749153137207, "num_tokens": 834308479.0, "step": 21864 }, { "epoch": 2.7814527413815036, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.483755111694336, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.882219672203064, "num_tokens": 834346185.0, "step": 21865 }, { "epoch": 2.781579951660094, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.364618301391602, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8821364641189575, "num_tokens": 834384404.0, "step": 21866 }, { "epoch": 2.7817071619386846, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.448722839355469, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8720459938049316, "num_tokens": 834424833.0, "step": 21867 }, { "epoch": 2.781834372217275, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.466975212097168, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8758341073989868, "num_tokens": 834464215.0, "step": 21868 }, { "epoch": 2.7819615824958657, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.467806816101074, "learning_rate": 1e-06, "loss": 0.4743, "mean_token_accuracy": 0.8510555028915405, "num_tokens": 834502452.0, "step": 21869 }, { "epoch": 2.782088792774456, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.487265586853027, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.873214066028595, "num_tokens": 834539690.0, "step": 21870 }, { "epoch": 2.7822160030530467, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.418832778930664, "learning_rate": 1e-06, "loss": 0.4337, "mean_token_accuracy": 0.8632652759552002, "num_tokens": 834583193.0, "step": 21871 }, { "epoch": 2.7823432133316373, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.409777641296387, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8677101135253906, "num_tokens": 834620028.0, "step": 21872 }, { "epoch": 2.782470423610228, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.440406799316406, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8636017441749573, "num_tokens": 834657351.0, "step": 21873 }, { "epoch": 2.7825976338888183, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.41358757019043, "learning_rate": 1e-06, "loss": 0.34, "mean_token_accuracy": 0.8883814811706543, "num_tokens": 834693687.0, "step": 21874 }, { "epoch": 2.782724844167409, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.421822547912598, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8722458481788635, "num_tokens": 834736171.0, "step": 21875 }, { "epoch": 2.7828520544459994, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.447509765625, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.886875569820404, "num_tokens": 834775699.0, "step": 21876 }, { "epoch": 2.78297926472459, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.452960968017578, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8675158023834229, "num_tokens": 834816053.0, "step": 21877 }, { "epoch": 2.78310647500318, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.492753028869629, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8701611757278442, "num_tokens": 834857125.0, "step": 21878 }, { "epoch": 2.783233685281771, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.448641777038574, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8862823247909546, "num_tokens": 834895676.0, "step": 21879 }, { "epoch": 2.783360895560361, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.494353294372559, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8766555786132812, "num_tokens": 834927748.0, "step": 21880 }, { "epoch": 2.783488105838952, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.497099876403809, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.869696855545044, "num_tokens": 834966722.0, "step": 21881 }, { "epoch": 2.783615316117542, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.430098533630371, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8752192854881287, "num_tokens": 835003770.0, "step": 21882 }, { "epoch": 2.783742526396133, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.470636367797852, "learning_rate": 1e-06, "loss": 0.375, "mean_token_accuracy": 0.8821114301681519, "num_tokens": 835041504.0, "step": 21883 }, { "epoch": 2.783869736674723, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.489706039428711, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8750892877578735, "num_tokens": 835075179.0, "step": 21884 }, { "epoch": 2.7839969469533137, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.48285961151123, "learning_rate": 1e-06, "loss": 0.4625, "mean_token_accuracy": 0.8630977272987366, "num_tokens": 835113080.0, "step": 21885 }, { "epoch": 2.784124157231904, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.454121589660645, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8723132610321045, "num_tokens": 835150586.0, "step": 21886 }, { "epoch": 2.7842513675104947, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.497881889343262, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8713517189025879, "num_tokens": 835184049.0, "step": 21887 }, { "epoch": 2.7843785777890853, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.460268020629883, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8684968948364258, "num_tokens": 835222745.0, "step": 21888 }, { "epoch": 2.784505788067676, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.452247619628906, "learning_rate": 1e-06, "loss": 0.4306, "mean_token_accuracy": 0.8611541986465454, "num_tokens": 835263017.0, "step": 21889 }, { "epoch": 2.7846329983462663, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.44365119934082, "learning_rate": 1e-06, "loss": 0.3568, "mean_token_accuracy": 0.8871928453445435, "num_tokens": 835303916.0, "step": 21890 }, { "epoch": 2.784760208624857, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.476699829101562, "learning_rate": 1e-06, "loss": 0.3614, "mean_token_accuracy": 0.8885326981544495, "num_tokens": 835343148.0, "step": 21891 }, { "epoch": 2.7848874189034474, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.501691818237305, "learning_rate": 1e-06, "loss": 0.4115, "mean_token_accuracy": 0.8712234497070312, "num_tokens": 835382337.0, "step": 21892 }, { "epoch": 2.785014629182038, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.573441505432129, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8667474985122681, "num_tokens": 835423645.0, "step": 21893 }, { "epoch": 2.7851418394606284, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45423698425293, "learning_rate": 1e-06, "loss": 0.4775, "mean_token_accuracy": 0.8466501235961914, "num_tokens": 835464590.0, "step": 21894 }, { "epoch": 2.785269049739219, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.561698913574219, "learning_rate": 1e-06, "loss": 0.3336, "mean_token_accuracy": 0.8951101899147034, "num_tokens": 835496376.0, "step": 21895 }, { "epoch": 2.7853962600178095, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.443465232849121, "learning_rate": 1e-06, "loss": 0.45, "mean_token_accuracy": 0.8586117029190063, "num_tokens": 835541231.0, "step": 21896 }, { "epoch": 2.7855234702964, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.547647476196289, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.8774639368057251, "num_tokens": 835584038.0, "step": 21897 }, { "epoch": 2.7856506805749905, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.480443954467773, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8670196533203125, "num_tokens": 835627189.0, "step": 21898 }, { "epoch": 2.785777890853581, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.526290893554688, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8744513988494873, "num_tokens": 835659701.0, "step": 21899 }, { "epoch": 2.7859051011321716, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.532123565673828, "learning_rate": 1e-06, "loss": 0.3676, "mean_token_accuracy": 0.8790744543075562, "num_tokens": 835694045.0, "step": 21900 }, { "epoch": 2.786032311410762, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.42458438873291, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8712204098701477, "num_tokens": 835732118.0, "step": 21901 }, { "epoch": 2.7861595216893527, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.532222747802734, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8822146654129028, "num_tokens": 835772700.0, "step": 21902 }, { "epoch": 2.7862867319679427, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.426941871643066, "learning_rate": 1e-06, "loss": 0.4624, "mean_token_accuracy": 0.8574563264846802, "num_tokens": 835811206.0, "step": 21903 }, { "epoch": 2.7864139422465337, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.50156307220459, "learning_rate": 1e-06, "loss": 0.3828, "mean_token_accuracy": 0.8786987066268921, "num_tokens": 835844381.0, "step": 21904 }, { "epoch": 2.786541152525124, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.490123748779297, "learning_rate": 1e-06, "loss": 0.424, "mean_token_accuracy": 0.8661108016967773, "num_tokens": 835879926.0, "step": 21905 }, { "epoch": 2.7866683628037148, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.386603355407715, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8625446557998657, "num_tokens": 835924014.0, "step": 21906 }, { "epoch": 2.786795573082305, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.455187797546387, "learning_rate": 1e-06, "loss": 0.4131, "mean_token_accuracy": 0.8665774464607239, "num_tokens": 835956054.0, "step": 21907 }, { "epoch": 2.786922783360896, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.514775276184082, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8796834945678711, "num_tokens": 835988885.0, "step": 21908 }, { "epoch": 2.787049993639486, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.412333488464355, "learning_rate": 1e-06, "loss": 0.4082, "mean_token_accuracy": 0.8725757598876953, "num_tokens": 836028885.0, "step": 21909 }, { "epoch": 2.7871772039180764, "ewc_loss": 0.0390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 14.438272476196289, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8869376182556152, "num_tokens": 836067340.0, "step": 21910 }, { "epoch": 2.787304414196667, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.502079963684082, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8736773729324341, "num_tokens": 836104432.0, "step": 21911 }, { "epoch": 2.7874316244752575, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.448244094848633, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.869204044342041, "num_tokens": 836141197.0, "step": 21912 }, { "epoch": 2.787558834753848, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.481596946716309, "learning_rate": 1e-06, "loss": 0.3691, "mean_token_accuracy": 0.8855277299880981, "num_tokens": 836176949.0, "step": 21913 }, { "epoch": 2.7876860450324386, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.47431468963623, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8709633350372314, "num_tokens": 836206484.0, "step": 21914 }, { "epoch": 2.787813255311029, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.489428520202637, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8787262439727783, "num_tokens": 836243056.0, "step": 21915 }, { "epoch": 2.7879404655896196, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.47952938079834, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8682811856269836, "num_tokens": 836281154.0, "step": 21916 }, { "epoch": 2.78806767586821, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.440388679504395, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8725234270095825, "num_tokens": 836319895.0, "step": 21917 }, { "epoch": 2.7881948861468007, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.458971977233887, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.878839373588562, "num_tokens": 836357227.0, "step": 21918 }, { "epoch": 2.788322096425391, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.406134605407715, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8664098381996155, "num_tokens": 836398442.0, "step": 21919 }, { "epoch": 2.7884493067039817, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.449268341064453, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8812252283096313, "num_tokens": 836436660.0, "step": 21920 }, { "epoch": 2.7885765169825723, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.481245994567871, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8677594661712646, "num_tokens": 836479180.0, "step": 21921 }, { "epoch": 2.788703727261163, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.409146308898926, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8794139623641968, "num_tokens": 836518874.0, "step": 21922 }, { "epoch": 2.7888309375397533, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.446205139160156, "learning_rate": 1e-06, "loss": 0.4094, "mean_token_accuracy": 0.8699651956558228, "num_tokens": 836556910.0, "step": 21923 }, { "epoch": 2.788958147818344, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.384203910827637, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8854498267173767, "num_tokens": 836598151.0, "step": 21924 }, { "epoch": 2.7890853580969344, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.42349624633789, "learning_rate": 1e-06, "loss": 0.4192, "mean_token_accuracy": 0.866401195526123, "num_tokens": 836637952.0, "step": 21925 }, { "epoch": 2.789212568375525, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.42722225189209, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8862301111221313, "num_tokens": 836677138.0, "step": 21926 }, { "epoch": 2.7893397786541154, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.384583473205566, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8780601620674133, "num_tokens": 836716880.0, "step": 21927 }, { "epoch": 2.7894669889327055, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.44864559173584, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8855887055397034, "num_tokens": 836752572.0, "step": 21928 }, { "epoch": 2.7895941992112965, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.38278579711914, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8694065809249878, "num_tokens": 836794321.0, "step": 21929 }, { "epoch": 2.7897214094898866, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.478217124938965, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.887498140335083, "num_tokens": 836829183.0, "step": 21930 }, { "epoch": 2.7898486197684775, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.342288970947266, "learning_rate": 1e-06, "loss": 0.4382, "mean_token_accuracy": 0.8624248504638672, "num_tokens": 836864644.0, "step": 21931 }, { "epoch": 2.7899758300470676, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.462689399719238, "learning_rate": 1e-06, "loss": 0.3657, "mean_token_accuracy": 0.8848550915718079, "num_tokens": 836895384.0, "step": 21932 }, { "epoch": 2.7901030403256586, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.379576683044434, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8764611482620239, "num_tokens": 836943336.0, "step": 21933 }, { "epoch": 2.7902302506042487, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.486186027526855, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8876341581344604, "num_tokens": 836980708.0, "step": 21934 }, { "epoch": 2.790357460882839, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.441569328308105, "learning_rate": 1e-06, "loss": 0.3934, "mean_token_accuracy": 0.8773773908615112, "num_tokens": 837015738.0, "step": 21935 }, { "epoch": 2.7904846711614297, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.421592712402344, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8802132606506348, "num_tokens": 837053014.0, "step": 21936 }, { "epoch": 2.7906118814400203, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.479982376098633, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8683196902275085, "num_tokens": 837083579.0, "step": 21937 }, { "epoch": 2.790739091718611, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.44903564453125, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8824578523635864, "num_tokens": 837126201.0, "step": 21938 }, { "epoch": 2.7908663019972013, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.472370147705078, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8824929594993591, "num_tokens": 837162400.0, "step": 21939 }, { "epoch": 2.790993512275792, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.429668426513672, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8777624368667603, "num_tokens": 837205733.0, "step": 21940 }, { "epoch": 2.7911207225543824, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.477434158325195, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8630849123001099, "num_tokens": 837243838.0, "step": 21941 }, { "epoch": 2.791247932832973, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.383975982666016, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8697556257247925, "num_tokens": 837280936.0, "step": 21942 }, { "epoch": 2.7913751431115634, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.46831226348877, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8737775087356567, "num_tokens": 837321790.0, "step": 21943 }, { "epoch": 2.791502353390154, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.540925979614258, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8769104480743408, "num_tokens": 837362023.0, "step": 21944 }, { "epoch": 2.7916295636687445, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.444421768188477, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8806363344192505, "num_tokens": 837404876.0, "step": 21945 }, { "epoch": 2.791756773947335, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.459419250488281, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8676251173019409, "num_tokens": 837442899.0, "step": 21946 }, { "epoch": 2.7918839842259255, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.461567878723145, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8839493989944458, "num_tokens": 837477611.0, "step": 21947 }, { "epoch": 2.792011194504516, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.44040298461914, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8612039685249329, "num_tokens": 837515928.0, "step": 21948 }, { "epoch": 2.7921384047831066, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.438645362854004, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8747271299362183, "num_tokens": 837553042.0, "step": 21949 }, { "epoch": 2.792265615061697, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.427661895751953, "learning_rate": 1e-06, "loss": 0.3858, "mean_token_accuracy": 0.8800216317176819, "num_tokens": 837589749.0, "step": 21950 }, { "epoch": 2.7923928253402877, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.44780445098877, "learning_rate": 1e-06, "loss": 0.3505, "mean_token_accuracy": 0.8921545743942261, "num_tokens": 837631104.0, "step": 21951 }, { "epoch": 2.792520035618878, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.458436965942383, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8835650086402893, "num_tokens": 837668173.0, "step": 21952 }, { "epoch": 2.7926472458974683, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.442309379577637, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8807798027992249, "num_tokens": 837704282.0, "step": 21953 }, { "epoch": 2.7927744561760592, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.462810516357422, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.864286482334137, "num_tokens": 837750790.0, "step": 21954 }, { "epoch": 2.7929016664546493, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.494341850280762, "learning_rate": 1e-06, "loss": 0.4013, "mean_token_accuracy": 0.8717639446258545, "num_tokens": 837792041.0, "step": 21955 }, { "epoch": 2.7930288767332403, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.478143692016602, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8803380727767944, "num_tokens": 837818645.0, "step": 21956 }, { "epoch": 2.7931560870118304, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.451372146606445, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8835703134536743, "num_tokens": 837859816.0, "step": 21957 }, { "epoch": 2.793283297290421, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.416533470153809, "learning_rate": 1e-06, "loss": 0.4169, "mean_token_accuracy": 0.8690703511238098, "num_tokens": 837903970.0, "step": 21958 }, { "epoch": 2.7934105075690114, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.49509048461914, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.874329149723053, "num_tokens": 837941158.0, "step": 21959 }, { "epoch": 2.793537717847602, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.444568634033203, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8808001279830933, "num_tokens": 837978157.0, "step": 21960 }, { "epoch": 2.7936649281261925, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.396770477294922, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8853743076324463, "num_tokens": 838017871.0, "step": 21961 }, { "epoch": 2.793792138404783, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.508295059204102, "learning_rate": 1e-06, "loss": 0.3625, "mean_token_accuracy": 0.8859862685203552, "num_tokens": 838050521.0, "step": 21962 }, { "epoch": 2.7939193486833735, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.460409164428711, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.882168710231781, "num_tokens": 838084450.0, "step": 21963 }, { "epoch": 2.794046558961964, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.46646785736084, "learning_rate": 1e-06, "loss": 0.4253, "mean_token_accuracy": 0.8689833283424377, "num_tokens": 838121505.0, "step": 21964 }, { "epoch": 2.7941737692405546, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.434990882873535, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8731203675270081, "num_tokens": 838166148.0, "step": 21965 }, { "epoch": 2.794300979519145, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.543474197387695, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8802918195724487, "num_tokens": 838203578.0, "step": 21966 }, { "epoch": 2.7944281897977357, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.458833694458008, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8796722888946533, "num_tokens": 838240702.0, "step": 21967 }, { "epoch": 2.794555400076326, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.48164176940918, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.88187175989151, "num_tokens": 838275639.0, "step": 21968 }, { "epoch": 2.7946826103549167, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.44769287109375, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8754643201828003, "num_tokens": 838310045.0, "step": 21969 }, { "epoch": 2.7948098206335072, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.4785737991333, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8751158714294434, "num_tokens": 838351571.0, "step": 21970 }, { "epoch": 2.7949370309120978, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.488349914550781, "learning_rate": 1e-06, "loss": 0.3609, "mean_token_accuracy": 0.8868886232376099, "num_tokens": 838393804.0, "step": 21971 }, { "epoch": 2.7950642411906883, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.456949234008789, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8736674785614014, "num_tokens": 838435516.0, "step": 21972 }, { "epoch": 2.795191451469279, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.498420715332031, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8816149234771729, "num_tokens": 838465425.0, "step": 21973 }, { "epoch": 2.7953186617478694, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.503297805786133, "learning_rate": 1e-06, "loss": 0.4077, "mean_token_accuracy": 0.8726007342338562, "num_tokens": 838512832.0, "step": 21974 }, { "epoch": 2.79544587202646, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.417132377624512, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8817882537841797, "num_tokens": 838552806.0, "step": 21975 }, { "epoch": 2.79557308230505, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.453760147094727, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8725699782371521, "num_tokens": 838588592.0, "step": 21976 }, { "epoch": 2.795700292583641, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.473876953125, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8729082345962524, "num_tokens": 838629402.0, "step": 21977 }, { "epoch": 2.795827502862231, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.421890258789062, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8661862015724182, "num_tokens": 838676580.0, "step": 21978 }, { "epoch": 2.795954713140822, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.435545921325684, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8695564270019531, "num_tokens": 838715773.0, "step": 21979 }, { "epoch": 2.796081923419412, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.409891128540039, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8639177083969116, "num_tokens": 838750270.0, "step": 21980 }, { "epoch": 2.796209133698003, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.406994819641113, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8536791801452637, "num_tokens": 838794823.0, "step": 21981 }, { "epoch": 2.796336343976593, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.470487594604492, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8664088249206543, "num_tokens": 838828901.0, "step": 21982 }, { "epoch": 2.7964635542551837, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.437891960144043, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8662692308425903, "num_tokens": 838872648.0, "step": 21983 }, { "epoch": 2.796590764533774, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.452536582946777, "learning_rate": 1e-06, "loss": 0.3571, "mean_token_accuracy": 0.8873429894447327, "num_tokens": 838905150.0, "step": 21984 }, { "epoch": 2.7967179748123647, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.464974403381348, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8747457265853882, "num_tokens": 838949027.0, "step": 21985 }, { "epoch": 2.7968451850909553, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.453195571899414, "learning_rate": 1e-06, "loss": 0.3931, "mean_token_accuracy": 0.8762087821960449, "num_tokens": 838987320.0, "step": 21986 }, { "epoch": 2.796972395369546, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.47356128692627, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.870483934879303, "num_tokens": 839027396.0, "step": 21987 }, { "epoch": 2.7970996056481363, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.415698051452637, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8800094127655029, "num_tokens": 839069779.0, "step": 21988 }, { "epoch": 2.797226815926727, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.422444343566895, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.881222128868103, "num_tokens": 839110919.0, "step": 21989 }, { "epoch": 2.7973540262053174, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.510002136230469, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8878709673881531, "num_tokens": 839143689.0, "step": 21990 }, { "epoch": 2.797481236483908, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.435325622558594, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8561782836914062, "num_tokens": 839181274.0, "step": 21991 }, { "epoch": 2.7976084467624984, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4893798828125, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8587490916252136, "num_tokens": 839217038.0, "step": 21992 }, { "epoch": 2.797735657041089, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.448620796203613, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8736156225204468, "num_tokens": 839260104.0, "step": 21993 }, { "epoch": 2.7978628673196795, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.472360610961914, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8776370286941528, "num_tokens": 839294328.0, "step": 21994 }, { "epoch": 2.79799007759827, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.439449310302734, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8730491399765015, "num_tokens": 839326662.0, "step": 21995 }, { "epoch": 2.7981172878768605, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.415689468383789, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.87578946352005, "num_tokens": 839355997.0, "step": 21996 }, { "epoch": 2.798244498155451, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.478102684020996, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8803675174713135, "num_tokens": 839396596.0, "step": 21997 }, { "epoch": 2.7983717084340416, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.46135139465332, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8757480382919312, "num_tokens": 839432251.0, "step": 21998 }, { "epoch": 2.798498918712632, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.45030403137207, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8836507797241211, "num_tokens": 839465987.0, "step": 21999 }, { "epoch": 2.7986261289912227, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.399669647216797, "learning_rate": 1e-06, "loss": 0.3947, "mean_token_accuracy": 0.8768922090530396, "num_tokens": 839506147.0, "step": 22000 }, { "epoch": 2.7987533392698127, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.481254577636719, "learning_rate": 1e-06, "loss": 0.4241, "mean_token_accuracy": 0.866841197013855, "num_tokens": 839544620.0, "step": 22001 }, { "epoch": 2.7988805495484037, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.441325187683105, "learning_rate": 1e-06, "loss": 0.4369, "mean_token_accuracy": 0.865947961807251, "num_tokens": 839583217.0, "step": 22002 }, { "epoch": 2.799007759826994, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.466414451599121, "learning_rate": 1e-06, "loss": 0.3528, "mean_token_accuracy": 0.8923275470733643, "num_tokens": 839621283.0, "step": 22003 }, { "epoch": 2.7991349701055848, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.469696044921875, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8701310157775879, "num_tokens": 839658072.0, "step": 22004 }, { "epoch": 2.799262180384175, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.508947372436523, "learning_rate": 1e-06, "loss": 0.4554, "mean_token_accuracy": 0.8573389053344727, "num_tokens": 839700926.0, "step": 22005 }, { "epoch": 2.799389390662766, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.477148056030273, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8833362460136414, "num_tokens": 839736160.0, "step": 22006 }, { "epoch": 2.799516600941356, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.452984809875488, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8891613483428955, "num_tokens": 839776834.0, "step": 22007 }, { "epoch": 2.7996438112199464, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.411381721496582, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8680688142776489, "num_tokens": 839818156.0, "step": 22008 }, { "epoch": 2.799771021498537, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.433976173400879, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8684513568878174, "num_tokens": 839857133.0, "step": 22009 }, { "epoch": 2.7998982317771275, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.471638679504395, "learning_rate": 1e-06, "loss": 0.4398, "mean_token_accuracy": 0.8593231439590454, "num_tokens": 839893129.0, "step": 22010 }, { "epoch": 2.800025442055718, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.445918083190918, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8776521682739258, "num_tokens": 839924622.0, "step": 22011 }, { "epoch": 2.8001526523343085, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45186996459961, "learning_rate": 1e-06, "loss": 0.3428, "mean_token_accuracy": 0.8935445547103882, "num_tokens": 839961276.0, "step": 22012 }, { "epoch": 2.800279862612899, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.442329406738281, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.8643411993980408, "num_tokens": 840002554.0, "step": 22013 }, { "epoch": 2.8004070728914896, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.40936279296875, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.873337984085083, "num_tokens": 840041642.0, "step": 22014 }, { "epoch": 2.80053428317008, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.446176528930664, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8829754590988159, "num_tokens": 840083532.0, "step": 22015 }, { "epoch": 2.8006614934486707, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.413958549499512, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8701891899108887, "num_tokens": 840124221.0, "step": 22016 }, { "epoch": 2.800788703727261, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.451944351196289, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8779051899909973, "num_tokens": 840162309.0, "step": 22017 }, { "epoch": 2.8009159140058517, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.417951583862305, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8758498430252075, "num_tokens": 840194778.0, "step": 22018 }, { "epoch": 2.8010431242844422, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.464863777160645, "learning_rate": 1e-06, "loss": 0.343, "mean_token_accuracy": 0.8923603892326355, "num_tokens": 840225081.0, "step": 22019 }, { "epoch": 2.8011703345630328, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.447071075439453, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8730728626251221, "num_tokens": 840267698.0, "step": 22020 }, { "epoch": 2.8012975448416233, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.453572273254395, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8670825362205505, "num_tokens": 840303255.0, "step": 22021 }, { "epoch": 2.801424755120214, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.48204231262207, "learning_rate": 1e-06, "loss": 0.3991, "mean_token_accuracy": 0.8735613226890564, "num_tokens": 840344616.0, "step": 22022 }, { "epoch": 2.8015519653988044, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.450796127319336, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8752008676528931, "num_tokens": 840383293.0, "step": 22023 }, { "epoch": 2.801679175677395, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.465646743774414, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8650057315826416, "num_tokens": 840424583.0, "step": 22024 }, { "epoch": 2.8018063859559854, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.460223197937012, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8687501549720764, "num_tokens": 840458752.0, "step": 22025 }, { "epoch": 2.8019335962345755, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.462380409240723, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8800422549247742, "num_tokens": 840498441.0, "step": 22026 }, { "epoch": 2.8020608065131665, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.469192504882812, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8839641809463501, "num_tokens": 840537838.0, "step": 22027 }, { "epoch": 2.8021880167917566, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.457420349121094, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8641238808631897, "num_tokens": 840571770.0, "step": 22028 }, { "epoch": 2.8023152270703475, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.433755874633789, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.8684342503547668, "num_tokens": 840607559.0, "step": 22029 }, { "epoch": 2.8024424373489376, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.48131275177002, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8600627183914185, "num_tokens": 840650156.0, "step": 22030 }, { "epoch": 2.8025696476275286, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.458227157592773, "learning_rate": 1e-06, "loss": 0.4349, "mean_token_accuracy": 0.8657410144805908, "num_tokens": 840686590.0, "step": 22031 }, { "epoch": 2.8026968579061187, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.448172569274902, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8666026592254639, "num_tokens": 840724445.0, "step": 22032 }, { "epoch": 2.802824068184709, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.44813346862793, "learning_rate": 1e-06, "loss": 0.3512, "mean_token_accuracy": 0.8887044191360474, "num_tokens": 840765047.0, "step": 22033 }, { "epoch": 2.8029512784632997, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.478670120239258, "learning_rate": 1e-06, "loss": 0.4359, "mean_token_accuracy": 0.8619830012321472, "num_tokens": 840802375.0, "step": 22034 }, { "epoch": 2.8030784887418903, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.472334861755371, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8736581206321716, "num_tokens": 840839994.0, "step": 22035 }, { "epoch": 2.803205699020481, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.433247566223145, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8726823925971985, "num_tokens": 840877144.0, "step": 22036 }, { "epoch": 2.8033329092990713, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.479917526245117, "learning_rate": 1e-06, "loss": 0.4322, "mean_token_accuracy": 0.8649460077285767, "num_tokens": 840918478.0, "step": 22037 }, { "epoch": 2.803460119577662, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.445110321044922, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8831605911254883, "num_tokens": 840958484.0, "step": 22038 }, { "epoch": 2.8035873298562524, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.457136154174805, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8765755891799927, "num_tokens": 841000035.0, "step": 22039 }, { "epoch": 2.803714540134843, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.46442985534668, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8689000606536865, "num_tokens": 841035836.0, "step": 22040 }, { "epoch": 2.8038417504134334, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.45982837677002, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8757380247116089, "num_tokens": 841069295.0, "step": 22041 }, { "epoch": 2.803968960692024, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.523898124694824, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8700357675552368, "num_tokens": 841108341.0, "step": 22042 }, { "epoch": 2.8040961709706145, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.461896896362305, "learning_rate": 1e-06, "loss": 0.3672, "mean_token_accuracy": 0.8837063312530518, "num_tokens": 841142945.0, "step": 22043 }, { "epoch": 2.804223381249205, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.518790245056152, "learning_rate": 1e-06, "loss": 0.3567, "mean_token_accuracy": 0.886805534362793, "num_tokens": 841174137.0, "step": 22044 }, { "epoch": 2.8043505915277955, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.43824291229248, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.8839753866195679, "num_tokens": 841215914.0, "step": 22045 }, { "epoch": 2.804477801806386, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.53222370147705, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8760483860969543, "num_tokens": 841256097.0, "step": 22046 }, { "epoch": 2.8046050120849766, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.471502304077148, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8777263164520264, "num_tokens": 841288997.0, "step": 22047 }, { "epoch": 2.804732222363567, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.515678405761719, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8689842820167542, "num_tokens": 841328053.0, "step": 22048 }, { "epoch": 2.8048594326421576, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.473011016845703, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8919066190719604, "num_tokens": 841366848.0, "step": 22049 }, { "epoch": 2.804986642920748, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.525871276855469, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8832961320877075, "num_tokens": 841398568.0, "step": 22050 }, { "epoch": 2.8051138531993383, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.50442123413086, "learning_rate": 1e-06, "loss": 0.4002, "mean_token_accuracy": 0.8737460970878601, "num_tokens": 841435722.0, "step": 22051 }, { "epoch": 2.8052410634779292, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.527202606201172, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8702741861343384, "num_tokens": 841473635.0, "step": 22052 }, { "epoch": 2.8053682737565193, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.470768928527832, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8771854639053345, "num_tokens": 841512383.0, "step": 22053 }, { "epoch": 2.8054954840351103, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.512313842773438, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8693889379501343, "num_tokens": 841543174.0, "step": 22054 }, { "epoch": 2.8056226943137004, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.467567443847656, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8790357112884521, "num_tokens": 841582185.0, "step": 22055 }, { "epoch": 2.805749904592291, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.50014877319336, "learning_rate": 1e-06, "loss": 0.4149, "mean_token_accuracy": 0.8687243461608887, "num_tokens": 841623179.0, "step": 22056 }, { "epoch": 2.8058771148708814, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.536773681640625, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8836988210678101, "num_tokens": 841661062.0, "step": 22057 }, { "epoch": 2.806004325149472, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.480920791625977, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8724371790885925, "num_tokens": 841704259.0, "step": 22058 }, { "epoch": 2.8061315354280625, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.5533447265625, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8716385364532471, "num_tokens": 841748396.0, "step": 22059 }, { "epoch": 2.806258745706653, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.530044555664062, "learning_rate": 1e-06, "loss": 0.3417, "mean_token_accuracy": 0.8923909068107605, "num_tokens": 841787809.0, "step": 22060 }, { "epoch": 2.8063859559852435, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.487698554992676, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8736469745635986, "num_tokens": 841824046.0, "step": 22061 }, { "epoch": 2.806513166263834, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.490082740783691, "learning_rate": 1e-06, "loss": 0.3637, "mean_token_accuracy": 0.884265661239624, "num_tokens": 841856949.0, "step": 22062 }, { "epoch": 2.8066403765424246, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.44202995300293, "learning_rate": 1e-06, "loss": 0.4469, "mean_token_accuracy": 0.8608139157295227, "num_tokens": 841898644.0, "step": 22063 }, { "epoch": 2.806767586821015, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.476234436035156, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8773767948150635, "num_tokens": 841934795.0, "step": 22064 }, { "epoch": 2.8068947970996057, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.45077133178711, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.858981728553772, "num_tokens": 841976878.0, "step": 22065 }, { "epoch": 2.807022007378196, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.443490028381348, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8746160864830017, "num_tokens": 842010998.0, "step": 22066 }, { "epoch": 2.8071492176567867, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.486760139465332, "learning_rate": 1e-06, "loss": 0.3227, "mean_token_accuracy": 0.8987369537353516, "num_tokens": 842051410.0, "step": 22067 }, { "epoch": 2.8072764279353772, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.462359428405762, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8764693737030029, "num_tokens": 842087954.0, "step": 22068 }, { "epoch": 2.8074036382139678, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.4815673828125, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8719454407691956, "num_tokens": 842128326.0, "step": 22069 }, { "epoch": 2.8075308484925583, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.485042572021484, "learning_rate": 1e-06, "loss": 0.468, "mean_token_accuracy": 0.8542997241020203, "num_tokens": 842171143.0, "step": 22070 }, { "epoch": 2.807658058771149, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.45683479309082, "learning_rate": 1e-06, "loss": 0.4334, "mean_token_accuracy": 0.8620855808258057, "num_tokens": 842212167.0, "step": 22071 }, { "epoch": 2.8077852690497394, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.500563621520996, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8646093606948853, "num_tokens": 842252576.0, "step": 22072 }, { "epoch": 2.80791247932833, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.469873428344727, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8702329993247986, "num_tokens": 842287812.0, "step": 22073 }, { "epoch": 2.80803968960692, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.477949142456055, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.870668888092041, "num_tokens": 842328289.0, "step": 22074 }, { "epoch": 2.808166899885511, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.51141357421875, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8777676820755005, "num_tokens": 842359182.0, "step": 22075 }, { "epoch": 2.808294110164101, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.448692321777344, "learning_rate": 1e-06, "loss": 0.4391, "mean_token_accuracy": 0.8640921711921692, "num_tokens": 842399820.0, "step": 22076 }, { "epoch": 2.808421320442692, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.50264835357666, "learning_rate": 1e-06, "loss": 0.4161, "mean_token_accuracy": 0.8724731206893921, "num_tokens": 842437852.0, "step": 22077 }, { "epoch": 2.808548530721282, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.401887893676758, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8609488010406494, "num_tokens": 842476729.0, "step": 22078 }, { "epoch": 2.808675740999873, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.459708213806152, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8624510765075684, "num_tokens": 842514590.0, "step": 22079 }, { "epoch": 2.808802951278463, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.549238204956055, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8726254105567932, "num_tokens": 842556502.0, "step": 22080 }, { "epoch": 2.8089301615570537, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.467352867126465, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.8630870580673218, "num_tokens": 842594299.0, "step": 22081 }, { "epoch": 2.809057371835644, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.466828346252441, "learning_rate": 1e-06, "loss": 0.4063, "mean_token_accuracy": 0.8709291815757751, "num_tokens": 842633189.0, "step": 22082 }, { "epoch": 2.8091845821142347, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.457071304321289, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8766528367996216, "num_tokens": 842673626.0, "step": 22083 }, { "epoch": 2.8093117923928252, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.52598762512207, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8670641779899597, "num_tokens": 842712574.0, "step": 22084 }, { "epoch": 2.8094390026714158, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.439935684204102, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8720716238021851, "num_tokens": 842749534.0, "step": 22085 }, { "epoch": 2.8095662129500063, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.527420043945312, "learning_rate": 1e-06, "loss": 0.428, "mean_token_accuracy": 0.8657149076461792, "num_tokens": 842784989.0, "step": 22086 }, { "epoch": 2.809693423228597, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.43029499053955, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8748921155929565, "num_tokens": 842823858.0, "step": 22087 }, { "epoch": 2.8098206335071874, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.509906768798828, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.8749328255653381, "num_tokens": 842858849.0, "step": 22088 }, { "epoch": 2.809947843785778, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.489082336425781, "learning_rate": 1e-06, "loss": 0.3408, "mean_token_accuracy": 0.893762469291687, "num_tokens": 842895809.0, "step": 22089 }, { "epoch": 2.8100750540643684, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.418254852294922, "learning_rate": 1e-06, "loss": 0.3163, "mean_token_accuracy": 0.9016081094741821, "num_tokens": 842933650.0, "step": 22090 }, { "epoch": 2.810202264342959, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.541192054748535, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8792102932929993, "num_tokens": 842967200.0, "step": 22091 }, { "epoch": 2.8103294746215495, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.44553279876709, "learning_rate": 1e-06, "loss": 0.4188, "mean_token_accuracy": 0.8656018376350403, "num_tokens": 843002498.0, "step": 22092 }, { "epoch": 2.81045668490014, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.450163841247559, "learning_rate": 1e-06, "loss": 0.3822, "mean_token_accuracy": 0.8789944052696228, "num_tokens": 843042627.0, "step": 22093 }, { "epoch": 2.8105838951787305, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.486141204833984, "learning_rate": 1e-06, "loss": 0.4009, "mean_token_accuracy": 0.8706446290016174, "num_tokens": 843076862.0, "step": 22094 }, { "epoch": 2.810711105457321, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.488299369812012, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8731340169906616, "num_tokens": 843115000.0, "step": 22095 }, { "epoch": 2.8108383157359116, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.468018531799316, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.8811063766479492, "num_tokens": 843152408.0, "step": 22096 }, { "epoch": 2.810965526014502, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.498507499694824, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8754358291625977, "num_tokens": 843189269.0, "step": 22097 }, { "epoch": 2.8110927362930926, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.469755172729492, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8703590035438538, "num_tokens": 843227148.0, "step": 22098 }, { "epoch": 2.8112199465716827, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.470333099365234, "learning_rate": 1e-06, "loss": 0.4251, "mean_token_accuracy": 0.8646188974380493, "num_tokens": 843259736.0, "step": 22099 }, { "epoch": 2.8113471568502737, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.527692794799805, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8700548410415649, "num_tokens": 843302598.0, "step": 22100 }, { "epoch": 2.811474367128864, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.467267036437988, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.888848066329956, "num_tokens": 843337615.0, "step": 22101 }, { "epoch": 2.8116015774074548, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.506911277770996, "learning_rate": 1e-06, "loss": 0.3575, "mean_token_accuracy": 0.889207661151886, "num_tokens": 843378845.0, "step": 22102 }, { "epoch": 2.811728787686045, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.461047172546387, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8643980026245117, "num_tokens": 843414128.0, "step": 22103 }, { "epoch": 2.811855997964636, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.478286743164062, "learning_rate": 1e-06, "loss": 0.3576, "mean_token_accuracy": 0.8892685174942017, "num_tokens": 843454255.0, "step": 22104 }, { "epoch": 2.811983208243226, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.437267303466797, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.875669002532959, "num_tokens": 843498332.0, "step": 22105 }, { "epoch": 2.8121104185218164, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.465140342712402, "learning_rate": 1e-06, "loss": 0.4312, "mean_token_accuracy": 0.8650776147842407, "num_tokens": 843539196.0, "step": 22106 }, { "epoch": 2.812237628800407, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.498273849487305, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.87822425365448, "num_tokens": 843573767.0, "step": 22107 }, { "epoch": 2.8123648390789975, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.443283081054688, "learning_rate": 1e-06, "loss": 0.4362, "mean_token_accuracy": 0.8628416061401367, "num_tokens": 843611156.0, "step": 22108 }, { "epoch": 2.812492049357588, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.552559852600098, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8803318738937378, "num_tokens": 843644026.0, "step": 22109 }, { "epoch": 2.8126192596361785, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.465068817138672, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8775200843811035, "num_tokens": 843683477.0, "step": 22110 }, { "epoch": 2.812746469914769, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.517670631408691, "learning_rate": 1e-06, "loss": 0.4418, "mean_token_accuracy": 0.8621558547019958, "num_tokens": 843722815.0, "step": 22111 }, { "epoch": 2.8128736801933596, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.47948932647705, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8758395910263062, "num_tokens": 843763843.0, "step": 22112 }, { "epoch": 2.81300089047195, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.505597114562988, "learning_rate": 1e-06, "loss": 0.4296, "mean_token_accuracy": 0.8666766881942749, "num_tokens": 843805327.0, "step": 22113 }, { "epoch": 2.8131281007505406, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.499550819396973, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.874910831451416, "num_tokens": 843850132.0, "step": 22114 }, { "epoch": 2.813255311029131, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.464068412780762, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8745315670967102, "num_tokens": 843887425.0, "step": 22115 }, { "epoch": 2.8133825213077217, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.449026107788086, "learning_rate": 1e-06, "loss": 0.3711, "mean_token_accuracy": 0.8825907707214355, "num_tokens": 843920728.0, "step": 22116 }, { "epoch": 2.8135097315863122, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.487922668457031, "learning_rate": 1e-06, "loss": 0.4023, "mean_token_accuracy": 0.873303234577179, "num_tokens": 843959733.0, "step": 22117 }, { "epoch": 2.8136369418649028, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.434687614440918, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8754863739013672, "num_tokens": 844005616.0, "step": 22118 }, { "epoch": 2.8137641521434933, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.441967964172363, "learning_rate": 1e-06, "loss": 0.3428, "mean_token_accuracy": 0.8885610699653625, "num_tokens": 844034216.0, "step": 22119 }, { "epoch": 2.813891362422084, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.521450996398926, "learning_rate": 1e-06, "loss": 0.4007, "mean_token_accuracy": 0.8710929155349731, "num_tokens": 844071173.0, "step": 22120 }, { "epoch": 2.8140185727006743, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.41734504699707, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8720724582672119, "num_tokens": 844113665.0, "step": 22121 }, { "epoch": 2.814145782979265, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.521836280822754, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8753554821014404, "num_tokens": 844154063.0, "step": 22122 }, { "epoch": 2.8142729932578554, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.441197395324707, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.8928751945495605, "num_tokens": 844185634.0, "step": 22123 }, { "epoch": 2.8144002035364455, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.527092933654785, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8647421598434448, "num_tokens": 844221755.0, "step": 22124 }, { "epoch": 2.8145274138150365, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.520694732666016, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8793326020240784, "num_tokens": 844256230.0, "step": 22125 }, { "epoch": 2.8146546240936265, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.555952072143555, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8681318759918213, "num_tokens": 844292823.0, "step": 22126 }, { "epoch": 2.8147818343722175, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.491063117980957, "learning_rate": 1e-06, "loss": 0.4661, "mean_token_accuracy": 0.8524504899978638, "num_tokens": 844331843.0, "step": 22127 }, { "epoch": 2.8149090446508076, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.490893363952637, "learning_rate": 1e-06, "loss": 0.372, "mean_token_accuracy": 0.8846216201782227, "num_tokens": 844375823.0, "step": 22128 }, { "epoch": 2.815036254929398, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.5103120803833, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8741127252578735, "num_tokens": 844415187.0, "step": 22129 }, { "epoch": 2.8151634652079887, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.465392112731934, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8769997358322144, "num_tokens": 844454986.0, "step": 22130 }, { "epoch": 2.815290675486579, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.56900691986084, "learning_rate": 1e-06, "loss": 0.3804, "mean_token_accuracy": 0.8803319334983826, "num_tokens": 844495514.0, "step": 22131 }, { "epoch": 2.8154178857651697, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.477490425109863, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8730993866920471, "num_tokens": 844532406.0, "step": 22132 }, { "epoch": 2.8155450960437602, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.503329277038574, "learning_rate": 1e-06, "loss": 0.4014, "mean_token_accuracy": 0.873187780380249, "num_tokens": 844569561.0, "step": 22133 }, { "epoch": 2.8156723063223508, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.504817962646484, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.879360020160675, "num_tokens": 844613504.0, "step": 22134 }, { "epoch": 2.8157995166009413, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.513802528381348, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8721383810043335, "num_tokens": 844649255.0, "step": 22135 }, { "epoch": 2.815926726879532, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.53724479675293, "learning_rate": 1e-06, "loss": 0.3976, "mean_token_accuracy": 0.8763954639434814, "num_tokens": 844687202.0, "step": 22136 }, { "epoch": 2.8160539371581224, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.490349769592285, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8755929470062256, "num_tokens": 844723096.0, "step": 22137 }, { "epoch": 2.816181147436713, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.470878601074219, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8760166764259338, "num_tokens": 844759754.0, "step": 22138 }, { "epoch": 2.8163083577153034, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.585082054138184, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.8784446716308594, "num_tokens": 844798372.0, "step": 22139 }, { "epoch": 2.816435567993894, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.553787231445312, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.8694596290588379, "num_tokens": 844832738.0, "step": 22140 }, { "epoch": 2.8165627782724845, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.525776863098145, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8762430548667908, "num_tokens": 844871684.0, "step": 22141 }, { "epoch": 2.816689988551075, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.48104190826416, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8671784996986389, "num_tokens": 844912140.0, "step": 22142 }, { "epoch": 2.8168171988296655, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.535353660583496, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.8630875945091248, "num_tokens": 844948256.0, "step": 22143 }, { "epoch": 2.816944409108256, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.5816068649292, "learning_rate": 1e-06, "loss": 0.3468, "mean_token_accuracy": 0.8902868628501892, "num_tokens": 844983821.0, "step": 22144 }, { "epoch": 2.8170716193868466, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.511798858642578, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8786061406135559, "num_tokens": 845020622.0, "step": 22145 }, { "epoch": 2.817198829665437, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.525425910949707, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8656213879585266, "num_tokens": 845056748.0, "step": 22146 }, { "epoch": 2.8173260399440276, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.557023048400879, "learning_rate": 1e-06, "loss": 0.4557, "mean_token_accuracy": 0.8565342426300049, "num_tokens": 845096294.0, "step": 22147 }, { "epoch": 2.817453250222618, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.518194198608398, "learning_rate": 1e-06, "loss": 0.3469, "mean_token_accuracy": 0.8898922204971313, "num_tokens": 845135132.0, "step": 22148 }, { "epoch": 2.8175804605012083, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.54792594909668, "learning_rate": 1e-06, "loss": 0.368, "mean_token_accuracy": 0.8829405307769775, "num_tokens": 845172164.0, "step": 22149 }, { "epoch": 2.8177076707797992, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.492417335510254, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8755671977996826, "num_tokens": 845210270.0, "step": 22150 }, { "epoch": 2.8178348810583893, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.457061767578125, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8779290914535522, "num_tokens": 845251863.0, "step": 22151 }, { "epoch": 2.8179620913369803, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.507250785827637, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.8791413307189941, "num_tokens": 845290262.0, "step": 22152 }, { "epoch": 2.8180893016155704, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.533724784851074, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8827857971191406, "num_tokens": 845326894.0, "step": 22153 }, { "epoch": 2.818216511894161, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.512240409851074, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.88337641954422, "num_tokens": 845367545.0, "step": 22154 }, { "epoch": 2.8183437221727514, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.518912315368652, "learning_rate": 1e-06, "loss": 0.4529, "mean_token_accuracy": 0.8604997396469116, "num_tokens": 845400920.0, "step": 22155 }, { "epoch": 2.818470932451342, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.546834945678711, "learning_rate": 1e-06, "loss": 0.4049, "mean_token_accuracy": 0.8714834451675415, "num_tokens": 845436279.0, "step": 22156 }, { "epoch": 2.8185981427299325, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.503013610839844, "learning_rate": 1e-06, "loss": 0.3834, "mean_token_accuracy": 0.8781729936599731, "num_tokens": 845476308.0, "step": 22157 }, { "epoch": 2.818725353008523, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.496552467346191, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.871019721031189, "num_tokens": 845516624.0, "step": 22158 }, { "epoch": 2.8188525632871135, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.52122688293457, "learning_rate": 1e-06, "loss": 0.3402, "mean_token_accuracy": 0.8923328518867493, "num_tokens": 845552385.0, "step": 22159 }, { "epoch": 2.818979773565704, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.499537467956543, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8841477036476135, "num_tokens": 845592600.0, "step": 22160 }, { "epoch": 2.8191069838442946, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.501473426818848, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8826776742935181, "num_tokens": 845630028.0, "step": 22161 }, { "epoch": 2.819234194122885, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.468927383422852, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8702949285507202, "num_tokens": 845665874.0, "step": 22162 }, { "epoch": 2.8193614044014756, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.533662796020508, "learning_rate": 1e-06, "loss": 0.435, "mean_token_accuracy": 0.8630936145782471, "num_tokens": 845702738.0, "step": 22163 }, { "epoch": 2.819488614680066, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.4934663772583, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8779484629631042, "num_tokens": 845744785.0, "step": 22164 }, { "epoch": 2.8196158249586567, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.515335083007812, "learning_rate": 1e-06, "loss": 0.4035, "mean_token_accuracy": 0.8743485808372498, "num_tokens": 845779196.0, "step": 22165 }, { "epoch": 2.8197430352372472, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.530879974365234, "learning_rate": 1e-06, "loss": 0.4224, "mean_token_accuracy": 0.8653509616851807, "num_tokens": 845819016.0, "step": 22166 }, { "epoch": 2.8198702455158378, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.482950210571289, "learning_rate": 1e-06, "loss": 0.3877, "mean_token_accuracy": 0.8760716915130615, "num_tokens": 845855014.0, "step": 22167 }, { "epoch": 2.8199974557944283, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.457401275634766, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8832080364227295, "num_tokens": 845895491.0, "step": 22168 }, { "epoch": 2.820124666073019, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.485360145568848, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8655503988265991, "num_tokens": 845931560.0, "step": 22169 }, { "epoch": 2.8202518763516093, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.487927436828613, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8748166561126709, "num_tokens": 845966701.0, "step": 22170 }, { "epoch": 2.8203790866302, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.405379295349121, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.891882061958313, "num_tokens": 846002327.0, "step": 22171 }, { "epoch": 2.82050629690879, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.439709663391113, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8749620914459229, "num_tokens": 846042157.0, "step": 22172 }, { "epoch": 2.820633507187381, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.488965034484863, "learning_rate": 1e-06, "loss": 0.4632, "mean_token_accuracy": 0.8597358465194702, "num_tokens": 846082851.0, "step": 22173 }, { "epoch": 2.820760717465971, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.461073875427246, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8736648559570312, "num_tokens": 846116846.0, "step": 22174 }, { "epoch": 2.820887927744562, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.47638988494873, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.877282440662384, "num_tokens": 846154373.0, "step": 22175 }, { "epoch": 2.821015138023152, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.435369491577148, "learning_rate": 1e-06, "loss": 0.3371, "mean_token_accuracy": 0.8924881219863892, "num_tokens": 846195329.0, "step": 22176 }, { "epoch": 2.821142348301743, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.494865417480469, "learning_rate": 1e-06, "loss": 0.3596, "mean_token_accuracy": 0.8868645429611206, "num_tokens": 846233078.0, "step": 22177 }, { "epoch": 2.821269558580333, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.491839408874512, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8737494945526123, "num_tokens": 846269030.0, "step": 22178 }, { "epoch": 2.8213967688589237, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.5131254196167, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8709592819213867, "num_tokens": 846307269.0, "step": 22179 }, { "epoch": 2.821523979137514, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.456377983093262, "learning_rate": 1e-06, "loss": 0.426, "mean_token_accuracy": 0.8656432628631592, "num_tokens": 846339116.0, "step": 22180 }, { "epoch": 2.8216511894161047, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.513792991638184, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8663526177406311, "num_tokens": 846384968.0, "step": 22181 }, { "epoch": 2.8217783996946952, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.510249137878418, "learning_rate": 1e-06, "loss": 0.4448, "mean_token_accuracy": 0.8593482971191406, "num_tokens": 846426235.0, "step": 22182 }, { "epoch": 2.8219056099732858, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.444613456726074, "learning_rate": 1e-06, "loss": 0.4351, "mean_token_accuracy": 0.8622294664382935, "num_tokens": 846465715.0, "step": 22183 }, { "epoch": 2.8220328202518763, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.540435791015625, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8614732623100281, "num_tokens": 846499972.0, "step": 22184 }, { "epoch": 2.822160030530467, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.486050605773926, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8730781078338623, "num_tokens": 846535906.0, "step": 22185 }, { "epoch": 2.8222872408090574, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.438836097717285, "learning_rate": 1e-06, "loss": 0.414, "mean_token_accuracy": 0.8704832792282104, "num_tokens": 846572723.0, "step": 22186 }, { "epoch": 2.822414451087648, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.571711540222168, "learning_rate": 1e-06, "loss": 0.3853, "mean_token_accuracy": 0.8792921304702759, "num_tokens": 846613501.0, "step": 22187 }, { "epoch": 2.8225416613662384, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.430201530456543, "learning_rate": 1e-06, "loss": 0.3423, "mean_token_accuracy": 0.8952183723449707, "num_tokens": 846650001.0, "step": 22188 }, { "epoch": 2.822668871644829, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.483266830444336, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.8632386922836304, "num_tokens": 846690636.0, "step": 22189 }, { "epoch": 2.8227960819234195, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.444384574890137, "learning_rate": 1e-06, "loss": 0.4683, "mean_token_accuracy": 0.8538900017738342, "num_tokens": 846728513.0, "step": 22190 }, { "epoch": 2.82292329220201, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.518606185913086, "learning_rate": 1e-06, "loss": 0.445, "mean_token_accuracy": 0.8603343963623047, "num_tokens": 846771989.0, "step": 22191 }, { "epoch": 2.8230505024806005, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.480525970458984, "learning_rate": 1e-06, "loss": 0.3453, "mean_token_accuracy": 0.8918131589889526, "num_tokens": 846808393.0, "step": 22192 }, { "epoch": 2.823177712759191, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.505278587341309, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8644428253173828, "num_tokens": 846846928.0, "step": 22193 }, { "epoch": 2.8233049230377816, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.433999061584473, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8654448390007019, "num_tokens": 846884637.0, "step": 22194 }, { "epoch": 2.823432133316372, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.532695770263672, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8777428865432739, "num_tokens": 846925749.0, "step": 22195 }, { "epoch": 2.8235593435949626, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.514034271240234, "learning_rate": 1e-06, "loss": 0.4524, "mean_token_accuracy": 0.8602734804153442, "num_tokens": 846967191.0, "step": 22196 }, { "epoch": 2.8236865538735527, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45194149017334, "learning_rate": 1e-06, "loss": 0.3778, "mean_token_accuracy": 0.8797448873519897, "num_tokens": 847003280.0, "step": 22197 }, { "epoch": 2.8238137641521437, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.537435531616211, "learning_rate": 1e-06, "loss": 0.4372, "mean_token_accuracy": 0.8620099425315857, "num_tokens": 847042179.0, "step": 22198 }, { "epoch": 2.8239409744307338, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.473274230957031, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.870163083076477, "num_tokens": 847083008.0, "step": 22199 }, { "epoch": 2.8240681847093247, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.54948616027832, "learning_rate": 1e-06, "loss": 0.4637, "mean_token_accuracy": 0.8535778522491455, "num_tokens": 847124009.0, "step": 22200 }, { "epoch": 2.824195394987915, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.535467147827148, "learning_rate": 1e-06, "loss": 0.4183, "mean_token_accuracy": 0.868335485458374, "num_tokens": 847163526.0, "step": 22201 }, { "epoch": 2.824322605266506, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.49821662902832, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8719978332519531, "num_tokens": 847201374.0, "step": 22202 }, { "epoch": 2.824449815545096, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.474472045898438, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8600152134895325, "num_tokens": 847239469.0, "step": 22203 }, { "epoch": 2.8245770258236864, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.533753395080566, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8720430135726929, "num_tokens": 847275964.0, "step": 22204 }, { "epoch": 2.824704236102277, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.447808265686035, "learning_rate": 1e-06, "loss": 0.4111, "mean_token_accuracy": 0.8755224943161011, "num_tokens": 847316815.0, "step": 22205 }, { "epoch": 2.8248314463808675, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.52097225189209, "learning_rate": 1e-06, "loss": 0.3866, "mean_token_accuracy": 0.8798366785049438, "num_tokens": 847350949.0, "step": 22206 }, { "epoch": 2.824958656659458, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.409198760986328, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8782273530960083, "num_tokens": 847388945.0, "step": 22207 }, { "epoch": 2.8250858669380485, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.520679473876953, "learning_rate": 1e-06, "loss": 0.3648, "mean_token_accuracy": 0.8852640390396118, "num_tokens": 847435114.0, "step": 22208 }, { "epoch": 2.825213077216639, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.435517311096191, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8653364181518555, "num_tokens": 847470815.0, "step": 22209 }, { "epoch": 2.8253402874952296, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.562561988830566, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8716686367988586, "num_tokens": 847509379.0, "step": 22210 }, { "epoch": 2.82546749777382, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.511248588562012, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8769847750663757, "num_tokens": 847547427.0, "step": 22211 }, { "epoch": 2.8255947080524106, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.479253768920898, "learning_rate": 1e-06, "loss": 0.4669, "mean_token_accuracy": 0.8570430278778076, "num_tokens": 847584740.0, "step": 22212 }, { "epoch": 2.825721918331001, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.529122352600098, "learning_rate": 1e-06, "loss": 0.4319, "mean_token_accuracy": 0.8622111678123474, "num_tokens": 847626201.0, "step": 22213 }, { "epoch": 2.8258491286095917, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.560016632080078, "learning_rate": 1e-06, "loss": 0.3588, "mean_token_accuracy": 0.8879585862159729, "num_tokens": 847661311.0, "step": 22214 }, { "epoch": 2.8259763388881822, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.487114906311035, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8715049028396606, "num_tokens": 847695234.0, "step": 22215 }, { "epoch": 2.8261035491667728, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.566970825195312, "learning_rate": 1e-06, "loss": 0.3439, "mean_token_accuracy": 0.890625, "num_tokens": 847738848.0, "step": 22216 }, { "epoch": 2.8262307594453633, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.560538291931152, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8769629001617432, "num_tokens": 847775952.0, "step": 22217 }, { "epoch": 2.826357969723954, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.52880859375, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8896229267120361, "num_tokens": 847809381.0, "step": 22218 }, { "epoch": 2.8264851800025443, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.505345344543457, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8701304197311401, "num_tokens": 847849835.0, "step": 22219 }, { "epoch": 2.826612390281135, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.494120597839355, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8719677925109863, "num_tokens": 847882726.0, "step": 22220 }, { "epoch": 2.8267396005597254, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.459840774536133, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.873623251914978, "num_tokens": 847920663.0, "step": 22221 }, { "epoch": 2.8268668108383155, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.558401107788086, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8810550570487976, "num_tokens": 847962204.0, "step": 22222 }, { "epoch": 2.8269940211169065, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.49435806274414, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8672840595245361, "num_tokens": 848003722.0, "step": 22223 }, { "epoch": 2.8271212313954965, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.498557090759277, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.8822600245475769, "num_tokens": 848045335.0, "step": 22224 }, { "epoch": 2.8272484416740875, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.53071403503418, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8661335706710815, "num_tokens": 848088817.0, "step": 22225 }, { "epoch": 2.8273756519526776, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.500540733337402, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8659230470657349, "num_tokens": 848120993.0, "step": 22226 }, { "epoch": 2.827502862231268, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.515888214111328, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8666614294052124, "num_tokens": 848162542.0, "step": 22227 }, { "epoch": 2.8276300725098586, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.467961311340332, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8875318765640259, "num_tokens": 848198194.0, "step": 22228 }, { "epoch": 2.827757282788449, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.528287887573242, "learning_rate": 1e-06, "loss": 0.4238, "mean_token_accuracy": 0.8684994578361511, "num_tokens": 848240575.0, "step": 22229 }, { "epoch": 2.8278844930670397, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.43696117401123, "learning_rate": 1e-06, "loss": 0.3986, "mean_token_accuracy": 0.8734720349311829, "num_tokens": 848289026.0, "step": 22230 }, { "epoch": 2.8280117033456302, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.504777908325195, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8757866621017456, "num_tokens": 848326438.0, "step": 22231 }, { "epoch": 2.8281389136242208, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.450159072875977, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8804130554199219, "num_tokens": 848368106.0, "step": 22232 }, { "epoch": 2.8282661239028113, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.469265937805176, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8705766797065735, "num_tokens": 848409775.0, "step": 22233 }, { "epoch": 2.828393334181402, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.499419212341309, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8740667104721069, "num_tokens": 848442149.0, "step": 22234 }, { "epoch": 2.8285205444599923, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.460846900939941, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8683803081512451, "num_tokens": 848475468.0, "step": 22235 }, { "epoch": 2.828647754738583, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.465682983398438, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8748147487640381, "num_tokens": 848509565.0, "step": 22236 }, { "epoch": 2.8287749650171734, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.469732284545898, "learning_rate": 1e-06, "loss": 0.3968, "mean_token_accuracy": 0.8752835988998413, "num_tokens": 848546357.0, "step": 22237 }, { "epoch": 2.828902175295764, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.517660140991211, "learning_rate": 1e-06, "loss": 0.4291, "mean_token_accuracy": 0.8624343872070312, "num_tokens": 848587661.0, "step": 22238 }, { "epoch": 2.8290293855743545, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.453783988952637, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8782773017883301, "num_tokens": 848628250.0, "step": 22239 }, { "epoch": 2.829156595852945, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.494001388549805, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8741860389709473, "num_tokens": 848670452.0, "step": 22240 }, { "epoch": 2.8292838061315355, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.510334014892578, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8756875991821289, "num_tokens": 848703592.0, "step": 22241 }, { "epoch": 2.829411016410126, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.534350395202637, "learning_rate": 1e-06, "loss": 0.4526, "mean_token_accuracy": 0.8577277064323425, "num_tokens": 848739291.0, "step": 22242 }, { "epoch": 2.8295382266887166, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.46978759765625, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8680911064147949, "num_tokens": 848776039.0, "step": 22243 }, { "epoch": 2.829665436967307, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.55599308013916, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8763371109962463, "num_tokens": 848814799.0, "step": 22244 }, { "epoch": 2.8297926472458976, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.527348518371582, "learning_rate": 1e-06, "loss": 0.3944, "mean_token_accuracy": 0.8751444816589355, "num_tokens": 848860123.0, "step": 22245 }, { "epoch": 2.829919857524488, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.485876083374023, "learning_rate": 1e-06, "loss": 0.3411, "mean_token_accuracy": 0.8906645178794861, "num_tokens": 848900447.0, "step": 22246 }, { "epoch": 2.8300470678030782, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.550321578979492, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8681970834732056, "num_tokens": 848945385.0, "step": 22247 }, { "epoch": 2.830174278081669, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.537256240844727, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8635900020599365, "num_tokens": 848978783.0, "step": 22248 }, { "epoch": 2.8303014883602593, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.600805282592773, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8624200224876404, "num_tokens": 849017324.0, "step": 22249 }, { "epoch": 2.8304286986388503, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.570436477661133, "learning_rate": 1e-06, "loss": 0.3445, "mean_token_accuracy": 0.893252968788147, "num_tokens": 849045928.0, "step": 22250 }, { "epoch": 2.8305559089174404, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.517642974853516, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8736275434494019, "num_tokens": 849080855.0, "step": 22251 }, { "epoch": 2.830683119196031, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.508190155029297, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8763942718505859, "num_tokens": 849114835.0, "step": 22252 }, { "epoch": 2.8308103294746214, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.522647857666016, "learning_rate": 1e-06, "loss": 0.4386, "mean_token_accuracy": 0.8633071780204773, "num_tokens": 849152632.0, "step": 22253 }, { "epoch": 2.830937539753212, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.50739574432373, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.882486879825592, "num_tokens": 849189911.0, "step": 22254 }, { "epoch": 2.8310647500318025, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.489728927612305, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8730465769767761, "num_tokens": 849227847.0, "step": 22255 }, { "epoch": 2.831191960310393, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.483083724975586, "learning_rate": 1e-06, "loss": 0.3721, "mean_token_accuracy": 0.8782262802124023, "num_tokens": 849265778.0, "step": 22256 }, { "epoch": 2.8313191705889835, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.527231216430664, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8734601736068726, "num_tokens": 849302488.0, "step": 22257 }, { "epoch": 2.831446380867574, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.534862518310547, "learning_rate": 1e-06, "loss": 0.3491, "mean_token_accuracy": 0.8890539407730103, "num_tokens": 849337483.0, "step": 22258 }, { "epoch": 2.8315735911461646, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.55095100402832, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8801013827323914, "num_tokens": 849378491.0, "step": 22259 }, { "epoch": 2.831700801424755, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.408498764038086, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8805027008056641, "num_tokens": 849418897.0, "step": 22260 }, { "epoch": 2.8318280117033456, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.526827812194824, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8785108923912048, "num_tokens": 849457057.0, "step": 22261 }, { "epoch": 2.831955221981936, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.509057998657227, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8778188228607178, "num_tokens": 849488825.0, "step": 22262 }, { "epoch": 2.8320824322605267, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.440101623535156, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8806215524673462, "num_tokens": 849528336.0, "step": 22263 }, { "epoch": 2.8322096425391172, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.547319412231445, "learning_rate": 1e-06, "loss": 0.3633, "mean_token_accuracy": 0.887291669845581, "num_tokens": 849565373.0, "step": 22264 }, { "epoch": 2.8323368528177078, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.52392292022705, "learning_rate": 1e-06, "loss": 0.3553, "mean_token_accuracy": 0.885206401348114, "num_tokens": 849600909.0, "step": 22265 }, { "epoch": 2.8324640630962983, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.509724617004395, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8654729127883911, "num_tokens": 849639499.0, "step": 22266 }, { "epoch": 2.832591273374889, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.506061553955078, "learning_rate": 1e-06, "loss": 0.3603, "mean_token_accuracy": 0.8857471346855164, "num_tokens": 849677232.0, "step": 22267 }, { "epoch": 2.8327184836534793, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.480632781982422, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8806055784225464, "num_tokens": 849717187.0, "step": 22268 }, { "epoch": 2.83284569393207, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.525222778320312, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8841554522514343, "num_tokens": 849756545.0, "step": 22269 }, { "epoch": 2.83297290421066, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.4848051071167, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8784331679344177, "num_tokens": 849788670.0, "step": 22270 }, { "epoch": 2.833100114489251, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.566600799560547, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8742692470550537, "num_tokens": 849825715.0, "step": 22271 }, { "epoch": 2.833227324767841, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.451772689819336, "learning_rate": 1e-06, "loss": 0.3763, "mean_token_accuracy": 0.8806031942367554, "num_tokens": 849856435.0, "step": 22272 }, { "epoch": 2.833354535046432, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.490848541259766, "learning_rate": 1e-06, "loss": 0.3538, "mean_token_accuracy": 0.8894588947296143, "num_tokens": 849895634.0, "step": 22273 }, { "epoch": 2.833481745325022, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.555561065673828, "learning_rate": 1e-06, "loss": 0.4589, "mean_token_accuracy": 0.8541370034217834, "num_tokens": 849938646.0, "step": 22274 }, { "epoch": 2.833608955603613, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.484916687011719, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8717513084411621, "num_tokens": 849978642.0, "step": 22275 }, { "epoch": 2.833736165882203, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.562488555908203, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8744813203811646, "num_tokens": 850017820.0, "step": 22276 }, { "epoch": 2.8338633761607936, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.52163314819336, "learning_rate": 1e-06, "loss": 0.4597, "mean_token_accuracy": 0.8558077812194824, "num_tokens": 850056619.0, "step": 22277 }, { "epoch": 2.833990586439384, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.55212116241455, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8722579479217529, "num_tokens": 850092456.0, "step": 22278 }, { "epoch": 2.8341177967179747, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.53634262084961, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8711104989051819, "num_tokens": 850131598.0, "step": 22279 }, { "epoch": 2.8342450069965652, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.480523109436035, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8647590279579163, "num_tokens": 850170884.0, "step": 22280 }, { "epoch": 2.8343722172751558, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.539419174194336, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8744539022445679, "num_tokens": 850206989.0, "step": 22281 }, { "epoch": 2.8344994275537463, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.478677749633789, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8706597089767456, "num_tokens": 850248602.0, "step": 22282 }, { "epoch": 2.834626637832337, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.570121765136719, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.866706132888794, "num_tokens": 850289790.0, "step": 22283 }, { "epoch": 2.8347538481109273, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.490126609802246, "learning_rate": 1e-06, "loss": 0.37, "mean_token_accuracy": 0.8831653594970703, "num_tokens": 850326688.0, "step": 22284 }, { "epoch": 2.834881058389518, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.491250038146973, "learning_rate": 1e-06, "loss": 0.3644, "mean_token_accuracy": 0.8857921361923218, "num_tokens": 850366460.0, "step": 22285 }, { "epoch": 2.8350082686681084, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.57235336303711, "learning_rate": 1e-06, "loss": 0.3765, "mean_token_accuracy": 0.8805707097053528, "num_tokens": 850400046.0, "step": 22286 }, { "epoch": 2.835135478946699, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.516420364379883, "learning_rate": 1e-06, "loss": 0.4376, "mean_token_accuracy": 0.8662525415420532, "num_tokens": 850434374.0, "step": 22287 }, { "epoch": 2.8352626892252895, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.522706031799316, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8759934902191162, "num_tokens": 850469056.0, "step": 22288 }, { "epoch": 2.83538989950388, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.533976554870605, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.871717095375061, "num_tokens": 850505805.0, "step": 22289 }, { "epoch": 2.8355171097824705, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.486358642578125, "learning_rate": 1e-06, "loss": 0.4388, "mean_token_accuracy": 0.8623126149177551, "num_tokens": 850550056.0, "step": 22290 }, { "epoch": 2.835644320061061, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.51648998260498, "learning_rate": 1e-06, "loss": 0.3813, "mean_token_accuracy": 0.881740927696228, "num_tokens": 850585075.0, "step": 22291 }, { "epoch": 2.8357715303396516, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.51823902130127, "learning_rate": 1e-06, "loss": 0.3671, "mean_token_accuracy": 0.8836908340454102, "num_tokens": 850625758.0, "step": 22292 }, { "epoch": 2.835898740618242, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.522592544555664, "learning_rate": 1e-06, "loss": 0.349, "mean_token_accuracy": 0.887654185295105, "num_tokens": 850664608.0, "step": 22293 }, { "epoch": 2.8360259508968326, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.531646728515625, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8820761442184448, "num_tokens": 850705897.0, "step": 22294 }, { "epoch": 2.8361531611754227, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.48099422454834, "learning_rate": 1e-06, "loss": 0.418, "mean_token_accuracy": 0.8694533109664917, "num_tokens": 850746195.0, "step": 22295 }, { "epoch": 2.8362803714540137, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.561660766601562, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8823880553245544, "num_tokens": 850785183.0, "step": 22296 }, { "epoch": 2.8364075817326038, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.475935935974121, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8806607723236084, "num_tokens": 850819558.0, "step": 22297 }, { "epoch": 2.8365347920111947, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.51325511932373, "learning_rate": 1e-06, "loss": 0.4287, "mean_token_accuracy": 0.8654136657714844, "num_tokens": 850857385.0, "step": 22298 }, { "epoch": 2.836662002289785, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.49647331237793, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.869918704032898, "num_tokens": 850895297.0, "step": 22299 }, { "epoch": 2.836789212568376, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.513605117797852, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8735859394073486, "num_tokens": 850930982.0, "step": 22300 }, { "epoch": 2.836916422846966, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.490586280822754, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8706598281860352, "num_tokens": 850969032.0, "step": 22301 }, { "epoch": 2.8370436331255564, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.531476974487305, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8782733082771301, "num_tokens": 851008325.0, "step": 22302 }, { "epoch": 2.837170843404147, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.558991432189941, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8765323162078857, "num_tokens": 851044289.0, "step": 22303 }, { "epoch": 2.8372980536827375, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.427252769470215, "learning_rate": 1e-06, "loss": 0.3775, "mean_token_accuracy": 0.8807222843170166, "num_tokens": 851079887.0, "step": 22304 }, { "epoch": 2.837425263961328, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.531902313232422, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8770081996917725, "num_tokens": 851124535.0, "step": 22305 }, { "epoch": 2.8375524742399185, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.538165092468262, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.883489727973938, "num_tokens": 851162982.0, "step": 22306 }, { "epoch": 2.837679684518509, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.532343864440918, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8817551136016846, "num_tokens": 851201864.0, "step": 22307 }, { "epoch": 2.8378068947970996, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.48922061920166, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8785653114318848, "num_tokens": 851241659.0, "step": 22308 }, { "epoch": 2.83793410507569, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.550921440124512, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8666627407073975, "num_tokens": 851278861.0, "step": 22309 }, { "epoch": 2.8380613153542806, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.514901161193848, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8763953447341919, "num_tokens": 851314442.0, "step": 22310 }, { "epoch": 2.838188525632871, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.50597095489502, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8697718381881714, "num_tokens": 851356322.0, "step": 22311 }, { "epoch": 2.8383157359114617, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.528402328491211, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8623993396759033, "num_tokens": 851395374.0, "step": 22312 }, { "epoch": 2.838442946190052, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.508432388305664, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8733124732971191, "num_tokens": 851430957.0, "step": 22313 }, { "epoch": 2.8385701564686427, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.51955509185791, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8697317838668823, "num_tokens": 851474357.0, "step": 22314 }, { "epoch": 2.8386973667472333, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.512364387512207, "learning_rate": 1e-06, "loss": 0.3552, "mean_token_accuracy": 0.8870325684547424, "num_tokens": 851510346.0, "step": 22315 }, { "epoch": 2.838824577025824, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.535082817077637, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8766912221908569, "num_tokens": 851546798.0, "step": 22316 }, { "epoch": 2.8389517873044143, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.517990112304688, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8804383277893066, "num_tokens": 851587819.0, "step": 22317 }, { "epoch": 2.839078997583005, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.597978591918945, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8791193962097168, "num_tokens": 851628764.0, "step": 22318 }, { "epoch": 2.8392062078615954, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.500260353088379, "learning_rate": 1e-06, "loss": 0.4387, "mean_token_accuracy": 0.8614372611045837, "num_tokens": 851672206.0, "step": 22319 }, { "epoch": 2.8393334181401855, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.544872283935547, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8690637350082397, "num_tokens": 851709969.0, "step": 22320 }, { "epoch": 2.8394606284187764, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.507731437683105, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.8685988187789917, "num_tokens": 851749015.0, "step": 22321 }, { "epoch": 2.8395878386973665, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.527822494506836, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8787004351615906, "num_tokens": 851787210.0, "step": 22322 }, { "epoch": 2.8397150489759575, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.519335746765137, "learning_rate": 1e-06, "loss": 0.4136, "mean_token_accuracy": 0.8677387833595276, "num_tokens": 851824677.0, "step": 22323 }, { "epoch": 2.8398422592545476, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.5823392868042, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.878476619720459, "num_tokens": 851863614.0, "step": 22324 }, { "epoch": 2.839969469533138, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.433788299560547, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.868013858795166, "num_tokens": 851908153.0, "step": 22325 }, { "epoch": 2.8400966798117286, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.604923248291016, "learning_rate": 1e-06, "loss": 0.4121, "mean_token_accuracy": 0.8716168403625488, "num_tokens": 851946473.0, "step": 22326 }, { "epoch": 2.840223890090319, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.510092735290527, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8638168573379517, "num_tokens": 851991600.0, "step": 22327 }, { "epoch": 2.8403511003689097, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.597777366638184, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.8762571215629578, "num_tokens": 852029468.0, "step": 22328 }, { "epoch": 2.8404783106475002, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.586397171020508, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8808459043502808, "num_tokens": 852061400.0, "step": 22329 }, { "epoch": 2.8406055209260908, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.577391624450684, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8740370273590088, "num_tokens": 852100596.0, "step": 22330 }, { "epoch": 2.8407327312046813, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.564984321594238, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8767929673194885, "num_tokens": 852140320.0, "step": 22331 }, { "epoch": 2.840859941483272, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.567683219909668, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.873479962348938, "num_tokens": 852175969.0, "step": 22332 }, { "epoch": 2.8409871517618623, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.568509101867676, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.8611083030700684, "num_tokens": 852210816.0, "step": 22333 }, { "epoch": 2.841114362040453, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.536231994628906, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8735562562942505, "num_tokens": 852246684.0, "step": 22334 }, { "epoch": 2.8412415723190434, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.546884536743164, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8773608207702637, "num_tokens": 852285283.0, "step": 22335 }, { "epoch": 2.841368782597634, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.553465843200684, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.8822906613349915, "num_tokens": 852323310.0, "step": 22336 }, { "epoch": 2.8414959928762245, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.540163040161133, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.878280758857727, "num_tokens": 852360850.0, "step": 22337 }, { "epoch": 2.841623203154815, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.53492259979248, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8679158687591553, "num_tokens": 852399320.0, "step": 22338 }, { "epoch": 2.8417504134334055, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.562817573547363, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8730398416519165, "num_tokens": 852430980.0, "step": 22339 }, { "epoch": 2.841877623711996, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.475772857666016, "learning_rate": 1e-06, "loss": 0.3701, "mean_token_accuracy": 0.8867419958114624, "num_tokens": 852465688.0, "step": 22340 }, { "epoch": 2.8420048339905866, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.580178260803223, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8730893135070801, "num_tokens": 852504372.0, "step": 22341 }, { "epoch": 2.842132044269177, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.470675468444824, "learning_rate": 1e-06, "loss": 0.3808, "mean_token_accuracy": 0.8789543509483337, "num_tokens": 852542232.0, "step": 22342 }, { "epoch": 2.8422592545477676, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.539133071899414, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.8654655814170837, "num_tokens": 852580005.0, "step": 22343 }, { "epoch": 2.842386464826358, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.468279838562012, "learning_rate": 1e-06, "loss": 0.3595, "mean_token_accuracy": 0.8865195512771606, "num_tokens": 852619912.0, "step": 22344 }, { "epoch": 2.8425136751049482, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.5545654296875, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8778307437896729, "num_tokens": 852654349.0, "step": 22345 }, { "epoch": 2.842640885383539, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.54219913482666, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8674543499946594, "num_tokens": 852691300.0, "step": 22346 }, { "epoch": 2.8427680956621293, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.525378227233887, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8731958866119385, "num_tokens": 852731228.0, "step": 22347 }, { "epoch": 2.8428953059407203, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.4859619140625, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8743138313293457, "num_tokens": 852768333.0, "step": 22348 }, { "epoch": 2.8430225162193103, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.542037963867188, "learning_rate": 1e-06, "loss": 0.495, "mean_token_accuracy": 0.8438957929611206, "num_tokens": 852810962.0, "step": 22349 }, { "epoch": 2.843149726497901, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.515551567077637, "learning_rate": 1e-06, "loss": 0.3551, "mean_token_accuracy": 0.8909413814544678, "num_tokens": 852844326.0, "step": 22350 }, { "epoch": 2.8432769367764914, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.571394920349121, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8689931631088257, "num_tokens": 852877737.0, "step": 22351 }, { "epoch": 2.843404147055082, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.537878036499023, "learning_rate": 1e-06, "loss": 0.3699, "mean_token_accuracy": 0.8820434212684631, "num_tokens": 852922085.0, "step": 22352 }, { "epoch": 2.8435313573336725, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.468374252319336, "learning_rate": 1e-06, "loss": 0.3424, "mean_token_accuracy": 0.8904169797897339, "num_tokens": 852960742.0, "step": 22353 }, { "epoch": 2.843658567612263, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.61168098449707, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8793613910675049, "num_tokens": 852993502.0, "step": 22354 }, { "epoch": 2.8437857778908535, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.480978012084961, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8788641691207886, "num_tokens": 853036844.0, "step": 22355 }, { "epoch": 2.843912988169444, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.57470417022705, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8734188079833984, "num_tokens": 853070851.0, "step": 22356 }, { "epoch": 2.8440401984480346, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.491358757019043, "learning_rate": 1e-06, "loss": 0.404, "mean_token_accuracy": 0.8749340772628784, "num_tokens": 853106116.0, "step": 22357 }, { "epoch": 2.844167408726625, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.448223114013672, "learning_rate": 1e-06, "loss": 0.402, "mean_token_accuracy": 0.8724607229232788, "num_tokens": 853152776.0, "step": 22358 }, { "epoch": 2.8442946190052156, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.57347297668457, "learning_rate": 1e-06, "loss": 0.3962, "mean_token_accuracy": 0.8728887438774109, "num_tokens": 853194176.0, "step": 22359 }, { "epoch": 2.844421829283806, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.61232852935791, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8724330067634583, "num_tokens": 853236834.0, "step": 22360 }, { "epoch": 2.8445490395623967, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.45910930633545, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.8666427135467529, "num_tokens": 853274145.0, "step": 22361 }, { "epoch": 2.844676249840987, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.567996978759766, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.877526044845581, "num_tokens": 853312569.0, "step": 22362 }, { "epoch": 2.8448034601195777, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.551254272460938, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8796921968460083, "num_tokens": 853351742.0, "step": 22363 }, { "epoch": 2.8449306703981683, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.549522399902344, "learning_rate": 1e-06, "loss": 0.3893, "mean_token_accuracy": 0.8766189813613892, "num_tokens": 853389607.0, "step": 22364 }, { "epoch": 2.845057880676759, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.540515899658203, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8814617395401001, "num_tokens": 853427218.0, "step": 22365 }, { "epoch": 2.8451850909553493, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.5737943649292, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8786497712135315, "num_tokens": 853464773.0, "step": 22366 }, { "epoch": 2.84531230123394, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.550992965698242, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8745012283325195, "num_tokens": 853507250.0, "step": 22367 }, { "epoch": 2.84543951151253, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.595158576965332, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8765435218811035, "num_tokens": 853551947.0, "step": 22368 }, { "epoch": 2.845566721791121, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.566909790039062, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8662616014480591, "num_tokens": 853590557.0, "step": 22369 }, { "epoch": 2.845693932069711, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.53364086151123, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8689269423484802, "num_tokens": 853630870.0, "step": 22370 }, { "epoch": 2.845821142348302, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.54687213897705, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8722507953643799, "num_tokens": 853662360.0, "step": 22371 }, { "epoch": 2.845948352626892, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.479597091674805, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8690588474273682, "num_tokens": 853703661.0, "step": 22372 }, { "epoch": 2.846075562905483, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.557191848754883, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8641993999481201, "num_tokens": 853742398.0, "step": 22373 }, { "epoch": 2.846202773184073, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.562396049499512, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8813282251358032, "num_tokens": 853781479.0, "step": 22374 }, { "epoch": 2.8463299834626636, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.532501220703125, "learning_rate": 1e-06, "loss": 0.4742, "mean_token_accuracy": 0.8497437238693237, "num_tokens": 853822330.0, "step": 22375 }, { "epoch": 2.846457193741254, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.507940292358398, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8625669479370117, "num_tokens": 853862747.0, "step": 22376 }, { "epoch": 2.8465844040198447, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.51977252960205, "learning_rate": 1e-06, "loss": 0.3282, "mean_token_accuracy": 0.897102952003479, "num_tokens": 853897124.0, "step": 22377 }, { "epoch": 2.8467116142984352, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.527323722839355, "learning_rate": 1e-06, "loss": 0.4453, "mean_token_accuracy": 0.8613337278366089, "num_tokens": 853930516.0, "step": 22378 }, { "epoch": 2.8468388245770258, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.57852840423584, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8787261247634888, "num_tokens": 853967016.0, "step": 22379 }, { "epoch": 2.8469660348556163, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.53853702545166, "learning_rate": 1e-06, "loss": 0.4394, "mean_token_accuracy": 0.8574711680412292, "num_tokens": 854008804.0, "step": 22380 }, { "epoch": 2.847093245134207, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.50578498840332, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.861653745174408, "num_tokens": 854048393.0, "step": 22381 }, { "epoch": 2.8472204554127973, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.567439079284668, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8791525363922119, "num_tokens": 854089480.0, "step": 22382 }, { "epoch": 2.847347665691388, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.498664855957031, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8845896124839783, "num_tokens": 854125677.0, "step": 22383 }, { "epoch": 2.8474748759699784, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.532623291015625, "learning_rate": 1e-06, "loss": 0.4146, "mean_token_accuracy": 0.8694714307785034, "num_tokens": 854160485.0, "step": 22384 }, { "epoch": 2.847602086248569, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.551004409790039, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8665754199028015, "num_tokens": 854200635.0, "step": 22385 }, { "epoch": 2.8477292965271594, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.548258781433105, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8786941766738892, "num_tokens": 854240046.0, "step": 22386 }, { "epoch": 2.84785650680575, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.51327896118164, "learning_rate": 1e-06, "loss": 0.3641, "mean_token_accuracy": 0.8848862648010254, "num_tokens": 854275880.0, "step": 22387 }, { "epoch": 2.8479837170843405, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.512585639953613, "learning_rate": 1e-06, "loss": 0.3956, "mean_token_accuracy": 0.8755846619606018, "num_tokens": 854311204.0, "step": 22388 }, { "epoch": 2.848110927362931, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.525091171264648, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.8714589476585388, "num_tokens": 854346076.0, "step": 22389 }, { "epoch": 2.8482381376415216, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.522939682006836, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8811870813369751, "num_tokens": 854383075.0, "step": 22390 }, { "epoch": 2.848365347920112, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.555522918701172, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8676639795303345, "num_tokens": 854416011.0, "step": 22391 }, { "epoch": 2.8484925581987026, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.486066818237305, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8764001131057739, "num_tokens": 854449520.0, "step": 22392 }, { "epoch": 2.8486197684772927, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.645914077758789, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8725007772445679, "num_tokens": 854484355.0, "step": 22393 }, { "epoch": 2.8487469787558837, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.592216491699219, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8605544567108154, "num_tokens": 854526238.0, "step": 22394 }, { "epoch": 2.8488741890344738, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.499101638793945, "learning_rate": 1e-06, "loss": 0.3856, "mean_token_accuracy": 0.8774822354316711, "num_tokens": 854562619.0, "step": 22395 }, { "epoch": 2.8490013993130647, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.522391319274902, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8795442581176758, "num_tokens": 854604974.0, "step": 22396 }, { "epoch": 2.849128609591655, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.587617874145508, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8806177377700806, "num_tokens": 854644066.0, "step": 22397 }, { "epoch": 2.849255819870246, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.56367301940918, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.87775057554245, "num_tokens": 854682723.0, "step": 22398 }, { "epoch": 2.849383030148836, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.549221992492676, "learning_rate": 1e-06, "loss": 0.3707, "mean_token_accuracy": 0.8850971460342407, "num_tokens": 854720292.0, "step": 22399 }, { "epoch": 2.8495102404274264, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.571256637573242, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8703542947769165, "num_tokens": 854756917.0, "step": 22400 }, { "epoch": 2.849637450706017, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.584274291992188, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8758426308631897, "num_tokens": 854798771.0, "step": 22401 }, { "epoch": 2.8497646609846075, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.518674850463867, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.8713129758834839, "num_tokens": 854838193.0, "step": 22402 }, { "epoch": 2.849891871263198, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.593297004699707, "learning_rate": 1e-06, "loss": 0.4218, "mean_token_accuracy": 0.8687918186187744, "num_tokens": 854871650.0, "step": 22403 }, { "epoch": 2.8500190815417885, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.564966201782227, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8728460073471069, "num_tokens": 854910757.0, "step": 22404 }, { "epoch": 2.850146291820379, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.6107816696167, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.8630435466766357, "num_tokens": 854945018.0, "step": 22405 }, { "epoch": 2.8502735020989696, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.554394721984863, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.866508424282074, "num_tokens": 854988369.0, "step": 22406 }, { "epoch": 2.85040071237756, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.564441680908203, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8765126466751099, "num_tokens": 855023935.0, "step": 22407 }, { "epoch": 2.8505279226561506, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.592071533203125, "learning_rate": 1e-06, "loss": 0.4069, "mean_token_accuracy": 0.8740633726119995, "num_tokens": 855062002.0, "step": 22408 }, { "epoch": 2.850655132934741, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.545496940612793, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8771089911460876, "num_tokens": 855101257.0, "step": 22409 }, { "epoch": 2.8507823432133317, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.532120704650879, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8763978481292725, "num_tokens": 855135834.0, "step": 22410 }, { "epoch": 2.850909553491922, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.569211959838867, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8764115571975708, "num_tokens": 855174156.0, "step": 22411 }, { "epoch": 2.8510367637705127, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.557024955749512, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8834787607192993, "num_tokens": 855207698.0, "step": 22412 }, { "epoch": 2.8511639740491033, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.555076599121094, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.8751283884048462, "num_tokens": 855240636.0, "step": 22413 }, { "epoch": 2.851291184327694, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.583718299865723, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8766935467720032, "num_tokens": 855279923.0, "step": 22414 }, { "epoch": 2.8514183946062843, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.52978515625, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8624171018600464, "num_tokens": 855315460.0, "step": 22415 }, { "epoch": 2.851545604884875, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.533260345458984, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8822089433670044, "num_tokens": 855351548.0, "step": 22416 }, { "epoch": 2.8516728151634654, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.51457405090332, "learning_rate": 1e-06, "loss": 0.4521, "mean_token_accuracy": 0.8567571640014648, "num_tokens": 855390715.0, "step": 22417 }, { "epoch": 2.8518000254420555, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.511454582214355, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8605127930641174, "num_tokens": 855431997.0, "step": 22418 }, { "epoch": 2.8519272357206464, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.53926944732666, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8676527738571167, "num_tokens": 855470242.0, "step": 22419 }, { "epoch": 2.8520544459992365, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.506549835205078, "learning_rate": 1e-06, "loss": 0.4321, "mean_token_accuracy": 0.8615154027938843, "num_tokens": 855508901.0, "step": 22420 }, { "epoch": 2.8521816562778275, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.58599853515625, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8700562119483948, "num_tokens": 855545290.0, "step": 22421 }, { "epoch": 2.8523088665564176, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.555115699768066, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8722692728042603, "num_tokens": 855579567.0, "step": 22422 }, { "epoch": 2.852436076835008, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.53492546081543, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8666092157363892, "num_tokens": 855621117.0, "step": 22423 }, { "epoch": 2.8525632871135986, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.602852821350098, "learning_rate": 1e-06, "loss": 0.3941, "mean_token_accuracy": 0.8765592575073242, "num_tokens": 855660494.0, "step": 22424 }, { "epoch": 2.852690497392189, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.565690040588379, "learning_rate": 1e-06, "loss": 0.4153, "mean_token_accuracy": 0.87370765209198, "num_tokens": 855694043.0, "step": 22425 }, { "epoch": 2.8528177076707797, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.558328628540039, "learning_rate": 1e-06, "loss": 0.4373, "mean_token_accuracy": 0.8594106435775757, "num_tokens": 855734615.0, "step": 22426 }, { "epoch": 2.85294491794937, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.541712760925293, "learning_rate": 1e-06, "loss": 0.3675, "mean_token_accuracy": 0.8843390941619873, "num_tokens": 855771141.0, "step": 22427 }, { "epoch": 2.8530721282279607, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.626230239868164, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8739051818847656, "num_tokens": 855806877.0, "step": 22428 }, { "epoch": 2.8531993385065513, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.55566120147705, "learning_rate": 1e-06, "loss": 0.3539, "mean_token_accuracy": 0.888978123664856, "num_tokens": 855840076.0, "step": 22429 }, { "epoch": 2.853326548785142, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.601140975952148, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8843112587928772, "num_tokens": 855877972.0, "step": 22430 }, { "epoch": 2.8534537590637323, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.582828521728516, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8689202070236206, "num_tokens": 855917424.0, "step": 22431 }, { "epoch": 2.853580969342323, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.559881210327148, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8637446165084839, "num_tokens": 855954853.0, "step": 22432 }, { "epoch": 2.8537081796209134, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.577543258666992, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.875411868095398, "num_tokens": 855993300.0, "step": 22433 }, { "epoch": 2.853835389899504, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.475852012634277, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8782147765159607, "num_tokens": 856035307.0, "step": 22434 }, { "epoch": 2.8539626001780944, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.53713607788086, "learning_rate": 1e-06, "loss": 0.4101, "mean_token_accuracy": 0.8728182911872864, "num_tokens": 856071801.0, "step": 22435 }, { "epoch": 2.854089810456685, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.515885353088379, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8621950149536133, "num_tokens": 856112035.0, "step": 22436 }, { "epoch": 2.8542170207352755, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.912882804870605, "learning_rate": 1e-06, "loss": 0.3843, "mean_token_accuracy": 0.8795523643493652, "num_tokens": 856148275.0, "step": 22437 }, { "epoch": 2.854344231013866, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.524383544921875, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.876899003982544, "num_tokens": 856180330.0, "step": 22438 }, { "epoch": 2.8544714412924566, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.426526069641113, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8618859052658081, "num_tokens": 856219354.0, "step": 22439 }, { "epoch": 2.854598651571047, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.530534744262695, "learning_rate": 1e-06, "loss": 0.3693, "mean_token_accuracy": 0.8840453624725342, "num_tokens": 856257583.0, "step": 22440 }, { "epoch": 2.8547258618496376, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.511923789978027, "learning_rate": 1e-06, "loss": 0.4465, "mean_token_accuracy": 0.8566436171531677, "num_tokens": 856298056.0, "step": 22441 }, { "epoch": 2.854853072128228, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.509302139282227, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8716821074485779, "num_tokens": 856334109.0, "step": 22442 }, { "epoch": 2.8549802824068182, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.533873558044434, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8760347366333008, "num_tokens": 856376655.0, "step": 22443 }, { "epoch": 2.855107492685409, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.47354507446289, "learning_rate": 1e-06, "loss": 0.3446, "mean_token_accuracy": 0.8900790214538574, "num_tokens": 856409926.0, "step": 22444 }, { "epoch": 2.8552347029639993, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.514988899230957, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8656553030014038, "num_tokens": 856451471.0, "step": 22445 }, { "epoch": 2.8553619132425903, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.55750560760498, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8825699090957642, "num_tokens": 856483237.0, "step": 22446 }, { "epoch": 2.8554891235211803, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.524833679199219, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8649567365646362, "num_tokens": 856521278.0, "step": 22447 }, { "epoch": 2.855616333799771, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.462154388427734, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8708937168121338, "num_tokens": 856565048.0, "step": 22448 }, { "epoch": 2.8557435440783614, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.551369667053223, "learning_rate": 1e-06, "loss": 0.3903, "mean_token_accuracy": 0.8768947124481201, "num_tokens": 856603756.0, "step": 22449 }, { "epoch": 2.855870754356952, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.521916389465332, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8798954486846924, "num_tokens": 856640677.0, "step": 22450 }, { "epoch": 2.8559979646355425, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.522143363952637, "learning_rate": 1e-06, "loss": 0.3911, "mean_token_accuracy": 0.8743181228637695, "num_tokens": 856679982.0, "step": 22451 }, { "epoch": 2.856125174914133, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.540462493896484, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8626199960708618, "num_tokens": 856720621.0, "step": 22452 }, { "epoch": 2.8562523851927235, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.558345794677734, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8826664686203003, "num_tokens": 856757172.0, "step": 22453 }, { "epoch": 2.856379595471314, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.52554702758789, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8707424998283386, "num_tokens": 856796876.0, "step": 22454 }, { "epoch": 2.8565068057499046, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.59936809539795, "learning_rate": 1e-06, "loss": 0.4011, "mean_token_accuracy": 0.8744025230407715, "num_tokens": 856831148.0, "step": 22455 }, { "epoch": 2.856634016028495, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.545374870300293, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8713840246200562, "num_tokens": 856867091.0, "step": 22456 }, { "epoch": 2.8567612263070856, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.589917182922363, "learning_rate": 1e-06, "loss": 0.4365, "mean_token_accuracy": 0.8648759126663208, "num_tokens": 856906118.0, "step": 22457 }, { "epoch": 2.856888436585676, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.536469459533691, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8487218618392944, "num_tokens": 856942694.0, "step": 22458 }, { "epoch": 2.8570156468642667, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.587859153747559, "learning_rate": 1e-06, "loss": 0.4211, "mean_token_accuracy": 0.8665391802787781, "num_tokens": 856977530.0, "step": 22459 }, { "epoch": 2.857142857142857, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.567096710205078, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8775641918182373, "num_tokens": 857012417.0, "step": 22460 }, { "epoch": 2.8572700674214477, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.492985725402832, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.8674456477165222, "num_tokens": 857053556.0, "step": 22461 }, { "epoch": 2.8573972777000383, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.53903579711914, "learning_rate": 1e-06, "loss": 0.3836, "mean_token_accuracy": 0.877194881439209, "num_tokens": 857094514.0, "step": 22462 }, { "epoch": 2.857524487978629, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.551789283752441, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8750113248825073, "num_tokens": 857133670.0, "step": 22463 }, { "epoch": 2.8576516982572193, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.537568092346191, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.8672271966934204, "num_tokens": 857167934.0, "step": 22464 }, { "epoch": 2.85777890853581, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.517333984375, "learning_rate": 1e-06, "loss": 0.451, "mean_token_accuracy": 0.8568049669265747, "num_tokens": 857208263.0, "step": 22465 }, { "epoch": 2.8579061188144, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.586270332336426, "learning_rate": 1e-06, "loss": 0.4553, "mean_token_accuracy": 0.8565763235092163, "num_tokens": 857255107.0, "step": 22466 }, { "epoch": 2.858033329092991, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.429713249206543, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.870987594127655, "num_tokens": 857298368.0, "step": 22467 }, { "epoch": 2.858160539371581, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.553668975830078, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8742942810058594, "num_tokens": 857336899.0, "step": 22468 }, { "epoch": 2.858287749650172, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.589555740356445, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8789151310920715, "num_tokens": 857378482.0, "step": 22469 }, { "epoch": 2.858414959928762, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.448007583618164, "learning_rate": 1e-06, "loss": 0.3722, "mean_token_accuracy": 0.883091926574707, "num_tokens": 857413203.0, "step": 22470 }, { "epoch": 2.858542170207353, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.570998191833496, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8718347549438477, "num_tokens": 857446706.0, "step": 22471 }, { "epoch": 2.858669380485943, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.547462463378906, "learning_rate": 1e-06, "loss": 0.4012, "mean_token_accuracy": 0.877686619758606, "num_tokens": 857483415.0, "step": 22472 }, { "epoch": 2.8587965907645336, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.559394836425781, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8784376978874207, "num_tokens": 857525323.0, "step": 22473 }, { "epoch": 2.858923801043124, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.56570053100586, "learning_rate": 1e-06, "loss": 0.3577, "mean_token_accuracy": 0.8910579681396484, "num_tokens": 857564400.0, "step": 22474 }, { "epoch": 2.8590510113217147, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.597707748413086, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8725957870483398, "num_tokens": 857602756.0, "step": 22475 }, { "epoch": 2.859178221600305, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.611600875854492, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8739131689071655, "num_tokens": 857636875.0, "step": 22476 }, { "epoch": 2.8593054318788957, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.520208358764648, "learning_rate": 1e-06, "loss": 0.3518, "mean_token_accuracy": 0.8904449939727783, "num_tokens": 857671950.0, "step": 22477 }, { "epoch": 2.8594326421574863, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.548124313354492, "learning_rate": 1e-06, "loss": 0.3629, "mean_token_accuracy": 0.886328399181366, "num_tokens": 857705274.0, "step": 22478 }, { "epoch": 2.859559852436077, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.569869995117188, "learning_rate": 1e-06, "loss": 0.4167, "mean_token_accuracy": 0.8680614233016968, "num_tokens": 857743730.0, "step": 22479 }, { "epoch": 2.8596870627146673, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.474691390991211, "learning_rate": 1e-06, "loss": 0.4039, "mean_token_accuracy": 0.8723055720329285, "num_tokens": 857784055.0, "step": 22480 }, { "epoch": 2.859814272993258, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.53993034362793, "learning_rate": 1e-06, "loss": 0.3741, "mean_token_accuracy": 0.8809804320335388, "num_tokens": 857822431.0, "step": 22481 }, { "epoch": 2.8599414832718484, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.539745330810547, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8855948448181152, "num_tokens": 857859258.0, "step": 22482 }, { "epoch": 2.860068693550439, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.615326881408691, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8658998012542725, "num_tokens": 857901209.0, "step": 22483 }, { "epoch": 2.8601959038290294, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.580497741699219, "learning_rate": 1e-06, "loss": 0.3578, "mean_token_accuracy": 0.8862783908843994, "num_tokens": 857935563.0, "step": 22484 }, { "epoch": 2.86032311410762, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.58310604095459, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8645291924476624, "num_tokens": 857966777.0, "step": 22485 }, { "epoch": 2.8604503243862105, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.643999099731445, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8794556260108948, "num_tokens": 858005078.0, "step": 22486 }, { "epoch": 2.860577534664801, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.56502628326416, "learning_rate": 1e-06, "loss": 0.4436, "mean_token_accuracy": 0.8618651628494263, "num_tokens": 858049288.0, "step": 22487 }, { "epoch": 2.8607047449433916, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.527764320373535, "learning_rate": 1e-06, "loss": 0.3414, "mean_token_accuracy": 0.894134521484375, "num_tokens": 858089834.0, "step": 22488 }, { "epoch": 2.860831955221982, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.557809829711914, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8786002397537231, "num_tokens": 858124152.0, "step": 22489 }, { "epoch": 2.8609591655005726, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.64217758178711, "learning_rate": 1e-06, "loss": 0.3869, "mean_token_accuracy": 0.8790850639343262, "num_tokens": 858163202.0, "step": 22490 }, { "epoch": 2.8610863757791627, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.532657623291016, "learning_rate": 1e-06, "loss": 0.3378, "mean_token_accuracy": 0.8928434252738953, "num_tokens": 858199173.0, "step": 22491 }, { "epoch": 2.8612135860577537, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.636870384216309, "learning_rate": 1e-06, "loss": 0.4092, "mean_token_accuracy": 0.8717663288116455, "num_tokens": 858239693.0, "step": 22492 }, { "epoch": 2.8613407963363438, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.553613662719727, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8814613819122314, "num_tokens": 858280720.0, "step": 22493 }, { "epoch": 2.8614680066149347, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.579779624938965, "learning_rate": 1e-06, "loss": 0.3618, "mean_token_accuracy": 0.8863511681556702, "num_tokens": 858314145.0, "step": 22494 }, { "epoch": 2.861595216893525, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.59276008605957, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8766521215438843, "num_tokens": 858355177.0, "step": 22495 }, { "epoch": 2.861722427172116, "ewc_loss": 0.03955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.48450756072998, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8785134553909302, "num_tokens": 858393070.0, "step": 22496 }, { "epoch": 2.861849637450706, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.58901309967041, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8742468357086182, "num_tokens": 858431827.0, "step": 22497 }, { "epoch": 2.8619768477292964, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.59683609008789, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8719791173934937, "num_tokens": 858474990.0, "step": 22498 }, { "epoch": 2.862104058007887, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.555232048034668, "learning_rate": 1e-06, "loss": 0.3615, "mean_token_accuracy": 0.8872005343437195, "num_tokens": 858514698.0, "step": 22499 }, { "epoch": 2.8622312682864774, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.603561401367188, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8757369518280029, "num_tokens": 858557601.0, "step": 22500 }, { "epoch": 2.862358478565068, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.571285247802734, "learning_rate": 1e-06, "loss": 0.3698, "mean_token_accuracy": 0.8829296827316284, "num_tokens": 858596830.0, "step": 22501 }, { "epoch": 2.8624856888436585, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.555513381958008, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8652366399765015, "num_tokens": 858637555.0, "step": 22502 }, { "epoch": 2.862612899122249, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.582329750061035, "learning_rate": 1e-06, "loss": 0.3368, "mean_token_accuracy": 0.8936929106712341, "num_tokens": 858672813.0, "step": 22503 }, { "epoch": 2.8627401094008396, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.569538116455078, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8802749514579773, "num_tokens": 858710743.0, "step": 22504 }, { "epoch": 2.86286731967943, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.619255065917969, "learning_rate": 1e-06, "loss": 0.3534, "mean_token_accuracy": 0.8872674703598022, "num_tokens": 858739486.0, "step": 22505 }, { "epoch": 2.8629945299580206, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.522845268249512, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8554028868675232, "num_tokens": 858780678.0, "step": 22506 }, { "epoch": 2.863121740236611, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.561999320983887, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8701156377792358, "num_tokens": 858826112.0, "step": 22507 }, { "epoch": 2.8632489505152017, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.560271263122559, "learning_rate": 1e-06, "loss": 0.3634, "mean_token_accuracy": 0.8853234648704529, "num_tokens": 858862158.0, "step": 22508 }, { "epoch": 2.863376160793792, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.57224178314209, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8794986605644226, "num_tokens": 858905270.0, "step": 22509 }, { "epoch": 2.8635033710723827, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.514814376831055, "learning_rate": 1e-06, "loss": 0.4085, "mean_token_accuracy": 0.8676881790161133, "num_tokens": 858941802.0, "step": 22510 }, { "epoch": 2.8636305813509733, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.600096702575684, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8778396248817444, "num_tokens": 858974672.0, "step": 22511 }, { "epoch": 2.863757791629564, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.496818542480469, "learning_rate": 1e-06, "loss": 0.4015, "mean_token_accuracy": 0.8735056519508362, "num_tokens": 859015305.0, "step": 22512 }, { "epoch": 2.8638850019081543, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.548026084899902, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8664770126342773, "num_tokens": 859050646.0, "step": 22513 }, { "epoch": 2.864012212186745, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.553632736206055, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8842307329177856, "num_tokens": 859090139.0, "step": 22514 }, { "epoch": 2.8641394224653354, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.52653980255127, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8792076110839844, "num_tokens": 859130661.0, "step": 22515 }, { "epoch": 2.8642666327439255, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.537236213684082, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8744663596153259, "num_tokens": 859166399.0, "step": 22516 }, { "epoch": 2.8643938430225164, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.55196475982666, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8748970031738281, "num_tokens": 859208349.0, "step": 22517 }, { "epoch": 2.8645210533011065, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.450787544250488, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8779075741767883, "num_tokens": 859243784.0, "step": 22518 }, { "epoch": 2.8646482635796975, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.595149993896484, "learning_rate": 1e-06, "loss": 0.4364, "mean_token_accuracy": 0.8612361550331116, "num_tokens": 859278345.0, "step": 22519 }, { "epoch": 2.8647754738582876, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.539758682250977, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8654513955116272, "num_tokens": 859317454.0, "step": 22520 }, { "epoch": 2.864902684136878, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.575657844543457, "learning_rate": 1e-06, "loss": 0.4143, "mean_token_accuracy": 0.8705243468284607, "num_tokens": 859352076.0, "step": 22521 }, { "epoch": 2.8650298944154686, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.547931671142578, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.882341206073761, "num_tokens": 859389957.0, "step": 22522 }, { "epoch": 2.865157104694059, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.58005142211914, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8744115233421326, "num_tokens": 859425167.0, "step": 22523 }, { "epoch": 2.8652843149726497, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.587088584899902, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8723076581954956, "num_tokens": 859470999.0, "step": 22524 }, { "epoch": 2.86541152525124, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.500465393066406, "learning_rate": 1e-06, "loss": 0.3717, "mean_token_accuracy": 0.8814581632614136, "num_tokens": 859503763.0, "step": 22525 }, { "epoch": 2.8655387355298307, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.538125038146973, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8829758167266846, "num_tokens": 859537739.0, "step": 22526 }, { "epoch": 2.8656659458084213, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.598909378051758, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8839037418365479, "num_tokens": 859573497.0, "step": 22527 }, { "epoch": 2.865793156087012, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.594290733337402, "learning_rate": 1e-06, "loss": 0.4102, "mean_token_accuracy": 0.8708966374397278, "num_tokens": 859606997.0, "step": 22528 }, { "epoch": 2.8659203663656023, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.63358211517334, "learning_rate": 1e-06, "loss": 0.4331, "mean_token_accuracy": 0.8631332516670227, "num_tokens": 859642278.0, "step": 22529 }, { "epoch": 2.866047576644193, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.508831977844238, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8857142925262451, "num_tokens": 859684728.0, "step": 22530 }, { "epoch": 2.8661747869227834, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.554421424865723, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.868366539478302, "num_tokens": 859720946.0, "step": 22531 }, { "epoch": 2.866301997201374, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.599455833435059, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8727930784225464, "num_tokens": 859752605.0, "step": 22532 }, { "epoch": 2.8664292074799644, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.605844497680664, "learning_rate": 1e-06, "loss": 0.3626, "mean_token_accuracy": 0.8871528506278992, "num_tokens": 859784206.0, "step": 22533 }, { "epoch": 2.866556417758555, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.621598243713379, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.870903491973877, "num_tokens": 859828142.0, "step": 22534 }, { "epoch": 2.8666836280371455, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.554526329040527, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8699262142181396, "num_tokens": 859870032.0, "step": 22535 }, { "epoch": 2.866810838315736, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.652661323547363, "learning_rate": 1e-06, "loss": 0.4247, "mean_token_accuracy": 0.865058422088623, "num_tokens": 859911417.0, "step": 22536 }, { "epoch": 2.8669380485943265, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.606159210205078, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8853895664215088, "num_tokens": 859953929.0, "step": 22537 }, { "epoch": 2.867065258872917, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.543423652648926, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8788051605224609, "num_tokens": 859985402.0, "step": 22538 }, { "epoch": 2.8671924691515076, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.6571683883667, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.872203528881073, "num_tokens": 860017374.0, "step": 22539 }, { "epoch": 2.867319679430098, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.566303253173828, "learning_rate": 1e-06, "loss": 0.366, "mean_token_accuracy": 0.884007453918457, "num_tokens": 860058217.0, "step": 22540 }, { "epoch": 2.867446889708688, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.606348037719727, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8589614033699036, "num_tokens": 860092317.0, "step": 22541 }, { "epoch": 2.867574099987279, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.662861824035645, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8709562420845032, "num_tokens": 860125067.0, "step": 22542 }, { "epoch": 2.8677013102658693, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.559450149536133, "learning_rate": 1e-06, "loss": 0.3608, "mean_token_accuracy": 0.8872714042663574, "num_tokens": 860163805.0, "step": 22543 }, { "epoch": 2.8678285205444602, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.563545227050781, "learning_rate": 1e-06, "loss": 0.4308, "mean_token_accuracy": 0.8620127439498901, "num_tokens": 860200178.0, "step": 22544 }, { "epoch": 2.8679557308230503, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.57856273651123, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8742230534553528, "num_tokens": 860235283.0, "step": 22545 }, { "epoch": 2.868082941101641, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.574869155883789, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.891678512096405, "num_tokens": 860276776.0, "step": 22546 }, { "epoch": 2.8682101513802314, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.574398040771484, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8793904781341553, "num_tokens": 860318492.0, "step": 22547 }, { "epoch": 2.868337361658822, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.600958824157715, "learning_rate": 1e-06, "loss": 0.433, "mean_token_accuracy": 0.8609616160392761, "num_tokens": 860358556.0, "step": 22548 }, { "epoch": 2.8684645719374124, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.585199356079102, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8657482862472534, "num_tokens": 860391114.0, "step": 22549 }, { "epoch": 2.868591782216003, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.618886947631836, "learning_rate": 1e-06, "loss": 0.4246, "mean_token_accuracy": 0.870180070400238, "num_tokens": 860430696.0, "step": 22550 }, { "epoch": 2.8687189924945935, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.513668060302734, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8819624185562134, "num_tokens": 860471055.0, "step": 22551 }, { "epoch": 2.868846202773184, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.534205436706543, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8682454228401184, "num_tokens": 860512472.0, "step": 22552 }, { "epoch": 2.8689734130517746, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.599878311157227, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8710479736328125, "num_tokens": 860541014.0, "step": 22553 }, { "epoch": 2.869100623330365, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.51195240020752, "learning_rate": 1e-06, "loss": 0.3545, "mean_token_accuracy": 0.8894205689430237, "num_tokens": 860585025.0, "step": 22554 }, { "epoch": 2.8692278336089556, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.624013900756836, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8832519054412842, "num_tokens": 860616361.0, "step": 22555 }, { "epoch": 2.869355043887546, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.514321327209473, "learning_rate": 1e-06, "loss": 0.4335, "mean_token_accuracy": 0.8666236400604248, "num_tokens": 860655976.0, "step": 22556 }, { "epoch": 2.8694822541661367, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.564925193786621, "learning_rate": 1e-06, "loss": 0.4357, "mean_token_accuracy": 0.8631916046142578, "num_tokens": 860703949.0, "step": 22557 }, { "epoch": 2.869609464444727, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.525259971618652, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8746547102928162, "num_tokens": 860739935.0, "step": 22558 }, { "epoch": 2.8697366747233177, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.55146312713623, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8780549764633179, "num_tokens": 860782483.0, "step": 22559 }, { "epoch": 2.8698638850019083, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.576034545898438, "learning_rate": 1e-06, "loss": 0.427, "mean_token_accuracy": 0.8637964129447937, "num_tokens": 860816821.0, "step": 22560 }, { "epoch": 2.869991095280499, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.581439018249512, "learning_rate": 1e-06, "loss": 0.4286, "mean_token_accuracy": 0.8627763986587524, "num_tokens": 860851313.0, "step": 22561 }, { "epoch": 2.8701183055590893, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.544573783874512, "learning_rate": 1e-06, "loss": 0.3725, "mean_token_accuracy": 0.8837100863456726, "num_tokens": 860890540.0, "step": 22562 }, { "epoch": 2.87024551583768, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.552777290344238, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.869380533695221, "num_tokens": 860929686.0, "step": 22563 }, { "epoch": 2.87037272611627, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.595609664916992, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8779228329658508, "num_tokens": 860968186.0, "step": 22564 }, { "epoch": 2.870499936394861, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.530418395996094, "learning_rate": 1e-06, "loss": 0.4064, "mean_token_accuracy": 0.8701250553131104, "num_tokens": 861007252.0, "step": 22565 }, { "epoch": 2.870627146673451, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.601632118225098, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.876067042350769, "num_tokens": 861049898.0, "step": 22566 }, { "epoch": 2.870754356952042, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.593770027160645, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8823779821395874, "num_tokens": 861087711.0, "step": 22567 }, { "epoch": 2.870881567230632, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.619316101074219, "learning_rate": 1e-06, "loss": 0.4582, "mean_token_accuracy": 0.8539729118347168, "num_tokens": 861128566.0, "step": 22568 }, { "epoch": 2.871008777509223, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.650968551635742, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8788363933563232, "num_tokens": 861156299.0, "step": 22569 }, { "epoch": 2.871135987787813, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.622679710388184, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8851869106292725, "num_tokens": 861189376.0, "step": 22570 }, { "epoch": 2.8712631980664036, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.641353607177734, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8715875148773193, "num_tokens": 861227156.0, "step": 22571 }, { "epoch": 2.871390408344994, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.617677688598633, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8741757273674011, "num_tokens": 861268088.0, "step": 22572 }, { "epoch": 2.8715176186235847, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.59638786315918, "learning_rate": 1e-06, "loss": 0.3865, "mean_token_accuracy": 0.8778348565101624, "num_tokens": 861308892.0, "step": 22573 }, { "epoch": 2.871644828902175, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.589858055114746, "learning_rate": 1e-06, "loss": 0.3888, "mean_token_accuracy": 0.8767017126083374, "num_tokens": 861350108.0, "step": 22574 }, { "epoch": 2.8717720391807657, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.600187301635742, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8847916126251221, "num_tokens": 861390553.0, "step": 22575 }, { "epoch": 2.8718992494593563, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.583497047424316, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8765084743499756, "num_tokens": 861425878.0, "step": 22576 }, { "epoch": 2.872026459737947, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.546929359436035, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8906452655792236, "num_tokens": 861460935.0, "step": 22577 }, { "epoch": 2.8721536700165373, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.650348663330078, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8592449426651001, "num_tokens": 861500446.0, "step": 22578 }, { "epoch": 2.872280880295128, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.546031951904297, "learning_rate": 1e-06, "loss": 0.4248, "mean_token_accuracy": 0.8633225560188293, "num_tokens": 861538000.0, "step": 22579 }, { "epoch": 2.8724080905737184, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.591545104980469, "learning_rate": 1e-06, "loss": 0.386, "mean_token_accuracy": 0.8776992559432983, "num_tokens": 861578052.0, "step": 22580 }, { "epoch": 2.872535300852309, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.651159286499023, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8705885410308838, "num_tokens": 861619065.0, "step": 22581 }, { "epoch": 2.8726625111308994, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.60657787322998, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8781914710998535, "num_tokens": 861660112.0, "step": 22582 }, { "epoch": 2.87278972140949, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.554462432861328, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.8765599131584167, "num_tokens": 861699312.0, "step": 22583 }, { "epoch": 2.8729169316880805, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.627344131469727, "learning_rate": 1e-06, "loss": 0.3434, "mean_token_accuracy": 0.8918214440345764, "num_tokens": 861735167.0, "step": 22584 }, { "epoch": 2.873044141966671, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.638407707214355, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8688215017318726, "num_tokens": 861778594.0, "step": 22585 }, { "epoch": 2.8731713522452615, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.582986831665039, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.874716579914093, "num_tokens": 861814754.0, "step": 22586 }, { "epoch": 2.873298562523852, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.589737892150879, "learning_rate": 1e-06, "loss": 0.4342, "mean_token_accuracy": 0.8608659505844116, "num_tokens": 861849019.0, "step": 22587 }, { "epoch": 2.8734257728024426, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.605778694152832, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8716599941253662, "num_tokens": 861889471.0, "step": 22588 }, { "epoch": 2.8735529830810327, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.513744354248047, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8658053874969482, "num_tokens": 861926811.0, "step": 22589 }, { "epoch": 2.8736801933596237, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.58618450164795, "learning_rate": 1e-06, "loss": 0.4262, "mean_token_accuracy": 0.8730793595314026, "num_tokens": 861963628.0, "step": 22590 }, { "epoch": 2.8738074036382137, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.622859001159668, "learning_rate": 1e-06, "loss": 0.4269, "mean_token_accuracy": 0.8635547161102295, "num_tokens": 862002915.0, "step": 22591 }, { "epoch": 2.8739346139168047, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.52777099609375, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.8809268474578857, "num_tokens": 862047276.0, "step": 22592 }, { "epoch": 2.874061824195395, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.557220458984375, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8824383616447449, "num_tokens": 862091241.0, "step": 22593 }, { "epoch": 2.8741890344739858, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.594894409179688, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8783590793609619, "num_tokens": 862126025.0, "step": 22594 }, { "epoch": 2.874316244752576, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.572604179382324, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8661088943481445, "num_tokens": 862160822.0, "step": 22595 }, { "epoch": 2.8744434550311664, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.580099105834961, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8790788650512695, "num_tokens": 862196808.0, "step": 22596 }, { "epoch": 2.874570665309757, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.586841583251953, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8605457544326782, "num_tokens": 862233722.0, "step": 22597 }, { "epoch": 2.8746978755883474, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.548554420471191, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8902815580368042, "num_tokens": 862271914.0, "step": 22598 }, { "epoch": 2.874825085866938, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.574018478393555, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8715099096298218, "num_tokens": 862313696.0, "step": 22599 }, { "epoch": 2.8749522961455285, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.689133644104004, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8683713674545288, "num_tokens": 862355366.0, "step": 22600 }, { "epoch": 2.875079506424119, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.582619667053223, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8659411072731018, "num_tokens": 862395630.0, "step": 22601 }, { "epoch": 2.8752067167027096, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.577997207641602, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8758169412612915, "num_tokens": 862429650.0, "step": 22602 }, { "epoch": 2.8753339269813, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.601907730102539, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8714668154716492, "num_tokens": 862470921.0, "step": 22603 }, { "epoch": 2.8754611372598906, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.611711502075195, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8808276653289795, "num_tokens": 862505204.0, "step": 22604 }, { "epoch": 2.875588347538481, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.577012062072754, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8736721277236938, "num_tokens": 862548686.0, "step": 22605 }, { "epoch": 2.8757155578170717, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.577232360839844, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8873138427734375, "num_tokens": 862585920.0, "step": 22606 }, { "epoch": 2.875842768095662, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.624972343444824, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.869940996170044, "num_tokens": 862619364.0, "step": 22607 }, { "epoch": 2.8759699783742527, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.655570983886719, "learning_rate": 1e-06, "loss": 0.4134, "mean_token_accuracy": 0.871523380279541, "num_tokens": 862661317.0, "step": 22608 }, { "epoch": 2.8760971886528433, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.627556800842285, "learning_rate": 1e-06, "loss": 0.3787, "mean_token_accuracy": 0.878481388092041, "num_tokens": 862698236.0, "step": 22609 }, { "epoch": 2.876224398931434, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.597058296203613, "learning_rate": 1e-06, "loss": 0.3619, "mean_token_accuracy": 0.884831428527832, "num_tokens": 862739224.0, "step": 22610 }, { "epoch": 2.8763516092100243, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.658856391906738, "learning_rate": 1e-06, "loss": 0.362, "mean_token_accuracy": 0.8889740705490112, "num_tokens": 862776534.0, "step": 22611 }, { "epoch": 2.876478819488615, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.63096809387207, "learning_rate": 1e-06, "loss": 0.4215, "mean_token_accuracy": 0.8678618669509888, "num_tokens": 862815690.0, "step": 22612 }, { "epoch": 2.8766060297672054, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.631340026855469, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.8733963966369629, "num_tokens": 862851021.0, "step": 22613 }, { "epoch": 2.8767332400457954, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.63766860961914, "learning_rate": 1e-06, "loss": 0.3738, "mean_token_accuracy": 0.8830389976501465, "num_tokens": 862885122.0, "step": 22614 }, { "epoch": 2.8768604503243864, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.544329643249512, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.8805875778198242, "num_tokens": 862922233.0, "step": 22615 }, { "epoch": 2.8769876606029765, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.672675132751465, "learning_rate": 1e-06, "loss": 0.367, "mean_token_accuracy": 0.8843069076538086, "num_tokens": 862961355.0, "step": 22616 }, { "epoch": 2.8771148708815675, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.664137840270996, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8832440972328186, "num_tokens": 862998523.0, "step": 22617 }, { "epoch": 2.8772420811601576, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.591882705688477, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8756693601608276, "num_tokens": 863039684.0, "step": 22618 }, { "epoch": 2.877369291438748, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.606392860412598, "learning_rate": 1e-06, "loss": 0.3849, "mean_token_accuracy": 0.8791841268539429, "num_tokens": 863077649.0, "step": 22619 }, { "epoch": 2.8774965017173386, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.59142017364502, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8701579570770264, "num_tokens": 863116816.0, "step": 22620 }, { "epoch": 2.877623711995929, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.596808433532715, "learning_rate": 1e-06, "loss": 0.3668, "mean_token_accuracy": 0.8850356340408325, "num_tokens": 863155306.0, "step": 22621 }, { "epoch": 2.8777509222745197, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.608136177062988, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8817589282989502, "num_tokens": 863194381.0, "step": 22622 }, { "epoch": 2.87787813255311, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.560120582580566, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8736987709999084, "num_tokens": 863231470.0, "step": 22623 }, { "epoch": 2.8780053428317007, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.64690113067627, "learning_rate": 1e-06, "loss": 0.4227, "mean_token_accuracy": 0.8677224516868591, "num_tokens": 863265617.0, "step": 22624 }, { "epoch": 2.8781325531102913, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.568842887878418, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8652822971343994, "num_tokens": 863300111.0, "step": 22625 }, { "epoch": 2.878259763388882, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.561836242675781, "learning_rate": 1e-06, "loss": 0.3679, "mean_token_accuracy": 0.8829665184020996, "num_tokens": 863338626.0, "step": 22626 }, { "epoch": 2.8783869736674723, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.549495697021484, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8675106763839722, "num_tokens": 863374672.0, "step": 22627 }, { "epoch": 2.878514183946063, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.600423812866211, "learning_rate": 1e-06, "loss": 0.4081, "mean_token_accuracy": 0.8699964880943298, "num_tokens": 863412717.0, "step": 22628 }, { "epoch": 2.8786413942246534, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.55853271484375, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8829742670059204, "num_tokens": 863447776.0, "step": 22629 }, { "epoch": 2.878768604503244, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.518034934997559, "learning_rate": 1e-06, "loss": 0.3484, "mean_token_accuracy": 0.891373336315155, "num_tokens": 863485697.0, "step": 22630 }, { "epoch": 2.8788958147818344, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.593841552734375, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8700200915336609, "num_tokens": 863528706.0, "step": 22631 }, { "epoch": 2.879023025060425, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.608915328979492, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8561916947364807, "num_tokens": 863569181.0, "step": 22632 }, { "epoch": 2.8791502353390155, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.641697883605957, "learning_rate": 1e-06, "loss": 0.4392, "mean_token_accuracy": 0.8596053123474121, "num_tokens": 863602985.0, "step": 22633 }, { "epoch": 2.879277445617606, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.600263595581055, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8795601725578308, "num_tokens": 863634590.0, "step": 22634 }, { "epoch": 2.8794046558961965, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.59812068939209, "learning_rate": 1e-06, "loss": 0.4158, "mean_token_accuracy": 0.8730462789535522, "num_tokens": 863677671.0, "step": 22635 }, { "epoch": 2.879531866174787, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.580277442932129, "learning_rate": 1e-06, "loss": 0.3952, "mean_token_accuracy": 0.8774113655090332, "num_tokens": 863716102.0, "step": 22636 }, { "epoch": 2.8796590764533776, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.573365211486816, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8668872117996216, "num_tokens": 863751560.0, "step": 22637 }, { "epoch": 2.879786286731968, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.625658988952637, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.8892602920532227, "num_tokens": 863792373.0, "step": 22638 }, { "epoch": 2.879913497010558, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.583236694335938, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8709188103675842, "num_tokens": 863829969.0, "step": 22639 }, { "epoch": 2.880040707289149, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.576732635498047, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8799593448638916, "num_tokens": 863867080.0, "step": 22640 }, { "epoch": 2.8801679175677393, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.542675018310547, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8786510229110718, "num_tokens": 863904230.0, "step": 22641 }, { "epoch": 2.8802951278463302, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.624480247497559, "learning_rate": 1e-06, "loss": 0.3883, "mean_token_accuracy": 0.8792279958724976, "num_tokens": 863947309.0, "step": 22642 }, { "epoch": 2.8804223381249203, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.602968215942383, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8800213932991028, "num_tokens": 863985557.0, "step": 22643 }, { "epoch": 2.880549548403511, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.592975616455078, "learning_rate": 1e-06, "loss": 0.4316, "mean_token_accuracy": 0.8636754751205444, "num_tokens": 864021842.0, "step": 22644 }, { "epoch": 2.8806767586821014, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.590482711791992, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.8677127361297607, "num_tokens": 864060765.0, "step": 22645 }, { "epoch": 2.880803968960692, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.547038078308105, "learning_rate": 1e-06, "loss": 0.3916, "mean_token_accuracy": 0.8778625130653381, "num_tokens": 864100624.0, "step": 22646 }, { "epoch": 2.8809311792392824, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.547405242919922, "learning_rate": 1e-06, "loss": 0.3635, "mean_token_accuracy": 0.8849376440048218, "num_tokens": 864136599.0, "step": 22647 }, { "epoch": 2.881058389517873, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.569616317749023, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8779833912849426, "num_tokens": 864169876.0, "step": 22648 }, { "epoch": 2.8811855997964635, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.517637252807617, "learning_rate": 1e-06, "loss": 0.3969, "mean_token_accuracy": 0.8736903667449951, "num_tokens": 864213709.0, "step": 22649 }, { "epoch": 2.881312810075054, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.624826431274414, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8804181814193726, "num_tokens": 864245581.0, "step": 22650 }, { "epoch": 2.8814400203536445, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.616887092590332, "learning_rate": 1e-06, "loss": 0.4482, "mean_token_accuracy": 0.8608382940292358, "num_tokens": 864287081.0, "step": 22651 }, { "epoch": 2.881567230632235, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.561739921569824, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8594622611999512, "num_tokens": 864321480.0, "step": 22652 }, { "epoch": 2.8816944409108256, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.578417778015137, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8816006779670715, "num_tokens": 864358580.0, "step": 22653 }, { "epoch": 2.881821651189416, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.550745964050293, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8657550811767578, "num_tokens": 864396277.0, "step": 22654 }, { "epoch": 2.8819488614680067, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.525300025939941, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.87065589427948, "num_tokens": 864433282.0, "step": 22655 }, { "epoch": 2.882076071746597, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.568053245544434, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.866740882396698, "num_tokens": 864469305.0, "step": 22656 }, { "epoch": 2.8822032820251877, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.557669639587402, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.8763221502304077, "num_tokens": 864507246.0, "step": 22657 }, { "epoch": 2.8823304923037782, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.548142433166504, "learning_rate": 1e-06, "loss": 0.4132, "mean_token_accuracy": 0.8688523769378662, "num_tokens": 864547315.0, "step": 22658 }, { "epoch": 2.8824577025823688, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.488988876342773, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8686295747756958, "num_tokens": 864590813.0, "step": 22659 }, { "epoch": 2.8825849128609593, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.592232704162598, "learning_rate": 1e-06, "loss": 0.4258, "mean_token_accuracy": 0.8659695982933044, "num_tokens": 864625078.0, "step": 22660 }, { "epoch": 2.88271212313955, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.49498176574707, "learning_rate": 1e-06, "loss": 0.3729, "mean_token_accuracy": 0.8825401067733765, "num_tokens": 864663714.0, "step": 22661 }, { "epoch": 2.88283933341814, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.603668212890625, "learning_rate": 1e-06, "loss": 0.4093, "mean_token_accuracy": 0.873742938041687, "num_tokens": 864697026.0, "step": 22662 }, { "epoch": 2.882966543696731, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.530868530273438, "learning_rate": 1e-06, "loss": 0.3753, "mean_token_accuracy": 0.8836420178413391, "num_tokens": 864732337.0, "step": 22663 }, { "epoch": 2.883093753975321, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.580277442932129, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8686388731002808, "num_tokens": 864776598.0, "step": 22664 }, { "epoch": 2.883220964253912, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.604397773742676, "learning_rate": 1e-06, "loss": 0.4456, "mean_token_accuracy": 0.8611448407173157, "num_tokens": 864817279.0, "step": 22665 }, { "epoch": 2.883348174532502, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.570338249206543, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8818389773368835, "num_tokens": 864854914.0, "step": 22666 }, { "epoch": 2.883475384811093, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.60676097869873, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8781387209892273, "num_tokens": 864892914.0, "step": 22667 }, { "epoch": 2.883602595089683, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.589668273925781, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8719345331192017, "num_tokens": 864934189.0, "step": 22668 }, { "epoch": 2.8837298053682736, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.571544647216797, "learning_rate": 1e-06, "loss": 0.3746, "mean_token_accuracy": 0.8842782974243164, "num_tokens": 864972085.0, "step": 22669 }, { "epoch": 2.883857015646864, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.578506469726562, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8584580421447754, "num_tokens": 865008674.0, "step": 22670 }, { "epoch": 2.8839842259254547, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.563292503356934, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8657391667366028, "num_tokens": 865046004.0, "step": 22671 }, { "epoch": 2.884111436204045, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.598556518554688, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8783649206161499, "num_tokens": 865082409.0, "step": 22672 }, { "epoch": 2.8842386464826357, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.519854545593262, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8740279674530029, "num_tokens": 865124457.0, "step": 22673 }, { "epoch": 2.8843658567612263, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.58080768585205, "learning_rate": 1e-06, "loss": 0.3682, "mean_token_accuracy": 0.8854316473007202, "num_tokens": 865162290.0, "step": 22674 }, { "epoch": 2.884493067039817, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.575797080993652, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8732342720031738, "num_tokens": 865201686.0, "step": 22675 }, { "epoch": 2.8846202773184073, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.615300178527832, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8652591705322266, "num_tokens": 865236876.0, "step": 22676 }, { "epoch": 2.884747487596998, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.592123985290527, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8779889345169067, "num_tokens": 865276994.0, "step": 22677 }, { "epoch": 2.8848746978755884, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.62229061126709, "learning_rate": 1e-06, "loss": 0.3885, "mean_token_accuracy": 0.873661994934082, "num_tokens": 865311990.0, "step": 22678 }, { "epoch": 2.885001908154179, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.53775405883789, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8806744813919067, "num_tokens": 865352324.0, "step": 22679 }, { "epoch": 2.8851291184327694, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.583800315856934, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8751609325408936, "num_tokens": 865391428.0, "step": 22680 }, { "epoch": 2.88525632871136, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.550985336303711, "learning_rate": 1e-06, "loss": 0.4016, "mean_token_accuracy": 0.8728970289230347, "num_tokens": 865428478.0, "step": 22681 }, { "epoch": 2.8853835389899505, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.635383605957031, "learning_rate": 1e-06, "loss": 0.4021, "mean_token_accuracy": 0.8750768899917603, "num_tokens": 865466531.0, "step": 22682 }, { "epoch": 2.885510749268541, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.594452857971191, "learning_rate": 1e-06, "loss": 0.3814, "mean_token_accuracy": 0.8810601830482483, "num_tokens": 865503177.0, "step": 22683 }, { "epoch": 2.8856379595471315, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.581997871398926, "learning_rate": 1e-06, "loss": 0.4216, "mean_token_accuracy": 0.8674108386039734, "num_tokens": 865539454.0, "step": 22684 }, { "epoch": 2.885765169825722, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.593538284301758, "learning_rate": 1e-06, "loss": 0.3454, "mean_token_accuracy": 0.8924959897994995, "num_tokens": 865571937.0, "step": 22685 }, { "epoch": 2.8858923801043126, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.599284172058105, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8767658472061157, "num_tokens": 865609457.0, "step": 22686 }, { "epoch": 2.8860195903829027, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.625323295593262, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8620089292526245, "num_tokens": 865647391.0, "step": 22687 }, { "epoch": 2.8861468006614936, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.587337493896484, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8733209371566772, "num_tokens": 865680578.0, "step": 22688 }, { "epoch": 2.8862740109400837, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.601068496704102, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8712809085845947, "num_tokens": 865714610.0, "step": 22689 }, { "epoch": 2.8864012212186747, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.6400728225708, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8701547384262085, "num_tokens": 865755672.0, "step": 22690 }, { "epoch": 2.886528431497265, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.605768203735352, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8822134733200073, "num_tokens": 865790111.0, "step": 22691 }, { "epoch": 2.8866556417758558, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.65799331665039, "learning_rate": 1e-06, "loss": 0.401, "mean_token_accuracy": 0.8752857446670532, "num_tokens": 865825580.0, "step": 22692 }, { "epoch": 2.886782852054446, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.589741706848145, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8773467540740967, "num_tokens": 865869099.0, "step": 22693 }, { "epoch": 2.8869100623330364, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.569766998291016, "learning_rate": 1e-06, "loss": 0.4428, "mean_token_accuracy": 0.8649095296859741, "num_tokens": 865906957.0, "step": 22694 }, { "epoch": 2.887037272611627, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.625164031982422, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8834932446479797, "num_tokens": 865945712.0, "step": 22695 }, { "epoch": 2.8871644828902174, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.554069519042969, "learning_rate": 1e-06, "loss": 0.4468, "mean_token_accuracy": 0.8605222702026367, "num_tokens": 865984469.0, "step": 22696 }, { "epoch": 2.887291693168808, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.629892349243164, "learning_rate": 1e-06, "loss": 0.403, "mean_token_accuracy": 0.8751142024993896, "num_tokens": 866024624.0, "step": 22697 }, { "epoch": 2.8874189034473985, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.549802780151367, "learning_rate": 1e-06, "loss": 0.4017, "mean_token_accuracy": 0.8713529109954834, "num_tokens": 866062446.0, "step": 22698 }, { "epoch": 2.887546113725989, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.611505508422852, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8836833238601685, "num_tokens": 866100896.0, "step": 22699 }, { "epoch": 2.8876733240045795, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.592957496643066, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8610411882400513, "num_tokens": 866134668.0, "step": 22700 }, { "epoch": 2.88780053428317, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.601629257202148, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.867342472076416, "num_tokens": 866173834.0, "step": 22701 }, { "epoch": 2.8879277445617606, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.598536491394043, "learning_rate": 1e-06, "loss": 0.4592, "mean_token_accuracy": 0.8582767844200134, "num_tokens": 866211691.0, "step": 22702 }, { "epoch": 2.888054954840351, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.645879745483398, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8738657236099243, "num_tokens": 866240878.0, "step": 22703 }, { "epoch": 2.8881821651189417, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.574651718139648, "learning_rate": 1e-06, "loss": 0.4347, "mean_token_accuracy": 0.8615294694900513, "num_tokens": 866289552.0, "step": 22704 }, { "epoch": 2.888309375397532, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.552353858947754, "learning_rate": 1e-06, "loss": 0.4328, "mean_token_accuracy": 0.8652034997940063, "num_tokens": 866330361.0, "step": 22705 }, { "epoch": 2.8884365856761227, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.585189819335938, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8778992295265198, "num_tokens": 866371557.0, "step": 22706 }, { "epoch": 2.8885637959547132, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.579283714294434, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8749948740005493, "num_tokens": 866410627.0, "step": 22707 }, { "epoch": 2.8886910062333038, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.637621879577637, "learning_rate": 1e-06, "loss": 0.4001, "mean_token_accuracy": 0.8732773065567017, "num_tokens": 866446036.0, "step": 22708 }, { "epoch": 2.8888182165118943, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.53537654876709, "learning_rate": 1e-06, "loss": 0.4288, "mean_token_accuracy": 0.8647439479827881, "num_tokens": 866489655.0, "step": 22709 }, { "epoch": 2.888945426790485, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.58364200592041, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8669165372848511, "num_tokens": 866532750.0, "step": 22710 }, { "epoch": 2.8890726370690754, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.646143913269043, "learning_rate": 1e-06, "loss": 0.3807, "mean_token_accuracy": 0.8804668188095093, "num_tokens": 866570748.0, "step": 22711 }, { "epoch": 2.8891998473476654, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.585087776184082, "learning_rate": 1e-06, "loss": 0.4184, "mean_token_accuracy": 0.869782567024231, "num_tokens": 866609422.0, "step": 22712 }, { "epoch": 2.8893270576262564, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.538650512695312, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8755340576171875, "num_tokens": 866650865.0, "step": 22713 }, { "epoch": 2.8894542679048465, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.541180610656738, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8723175525665283, "num_tokens": 866687605.0, "step": 22714 }, { "epoch": 2.8895814781834375, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.621838569641113, "learning_rate": 1e-06, "loss": 0.4232, "mean_token_accuracy": 0.8653988838195801, "num_tokens": 866722974.0, "step": 22715 }, { "epoch": 2.8897086884620276, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.58346176147461, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8717811107635498, "num_tokens": 866760694.0, "step": 22716 }, { "epoch": 2.889835898740618, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.623656272888184, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8570677042007446, "num_tokens": 866795102.0, "step": 22717 }, { "epoch": 2.8899631090192086, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.580619812011719, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8870824575424194, "num_tokens": 866831282.0, "step": 22718 }, { "epoch": 2.890090319297799, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.570382118225098, "learning_rate": 1e-06, "loss": 0.4231, "mean_token_accuracy": 0.8661706447601318, "num_tokens": 866872661.0, "step": 22719 }, { "epoch": 2.8902175295763897, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.571629524230957, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.86396723985672, "num_tokens": 866917576.0, "step": 22720 }, { "epoch": 2.89034473985498, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.58393669128418, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8776323199272156, "num_tokens": 866957245.0, "step": 22721 }, { "epoch": 2.8904719501335707, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.654129028320312, "learning_rate": 1e-06, "loss": 0.4773, "mean_token_accuracy": 0.8463471531867981, "num_tokens": 866998695.0, "step": 22722 }, { "epoch": 2.8905991604121613, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.660901069641113, "learning_rate": 1e-06, "loss": 0.4415, "mean_token_accuracy": 0.8608995079994202, "num_tokens": 867033502.0, "step": 22723 }, { "epoch": 2.890726370690752, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.562106132507324, "learning_rate": 1e-06, "loss": 0.4503, "mean_token_accuracy": 0.8567133545875549, "num_tokens": 867076032.0, "step": 22724 }, { "epoch": 2.8908535809693423, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.644919395446777, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.873818576335907, "num_tokens": 867107429.0, "step": 22725 }, { "epoch": 2.890980791247933, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.595672607421875, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8765202760696411, "num_tokens": 867139432.0, "step": 22726 }, { "epoch": 2.8911080015265234, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.59450912475586, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8707908391952515, "num_tokens": 867180588.0, "step": 22727 }, { "epoch": 2.891235211805114, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.625707626342773, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8710535764694214, "num_tokens": 867215170.0, "step": 22728 }, { "epoch": 2.8913624220837044, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.596202850341797, "learning_rate": 1e-06, "loss": 0.358, "mean_token_accuracy": 0.888556718826294, "num_tokens": 867254958.0, "step": 22729 }, { "epoch": 2.891489632362295, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.667826652526855, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8716765642166138, "num_tokens": 867291135.0, "step": 22730 }, { "epoch": 2.8916168426408855, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.617714881896973, "learning_rate": 1e-06, "loss": 0.3859, "mean_token_accuracy": 0.8817319869995117, "num_tokens": 867326926.0, "step": 22731 }, { "epoch": 2.891744052919476, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.607094764709473, "learning_rate": 1e-06, "loss": 0.4229, "mean_token_accuracy": 0.8657994270324707, "num_tokens": 867361142.0, "step": 22732 }, { "epoch": 2.8918712631980665, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.624560356140137, "learning_rate": 1e-06, "loss": 0.3509, "mean_token_accuracy": 0.8909726142883301, "num_tokens": 867398344.0, "step": 22733 }, { "epoch": 2.891998473476657, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.552741050720215, "learning_rate": 1e-06, "loss": 0.3758, "mean_token_accuracy": 0.8826380372047424, "num_tokens": 867437642.0, "step": 22734 }, { "epoch": 2.8921256837552476, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.68032455444336, "learning_rate": 1e-06, "loss": 0.4344, "mean_token_accuracy": 0.863767147064209, "num_tokens": 867471646.0, "step": 22735 }, { "epoch": 2.892252894033838, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.558039665222168, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8745723366737366, "num_tokens": 867509682.0, "step": 22736 }, { "epoch": 2.892380104312428, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.633152961730957, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.869631290435791, "num_tokens": 867544425.0, "step": 22737 }, { "epoch": 2.892507314591019, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.526545524597168, "learning_rate": 1e-06, "loss": 0.4208, "mean_token_accuracy": 0.8669625520706177, "num_tokens": 867584978.0, "step": 22738 }, { "epoch": 2.8926345248696093, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.56434154510498, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.874356746673584, "num_tokens": 867620360.0, "step": 22739 }, { "epoch": 2.8927617351482002, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.548176765441895, "learning_rate": 1e-06, "loss": 0.3684, "mean_token_accuracy": 0.8818281292915344, "num_tokens": 867655246.0, "step": 22740 }, { "epoch": 2.8928889454267903, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.574275016784668, "learning_rate": 1e-06, "loss": 0.4223, "mean_token_accuracy": 0.8653146624565125, "num_tokens": 867692792.0, "step": 22741 }, { "epoch": 2.893016155705381, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.568546295166016, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8766701221466064, "num_tokens": 867732365.0, "step": 22742 }, { "epoch": 2.8931433659839714, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.641120910644531, "learning_rate": 1e-06, "loss": 0.4443, "mean_token_accuracy": 0.8599602580070496, "num_tokens": 867764800.0, "step": 22743 }, { "epoch": 2.893270576262562, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.59574031829834, "learning_rate": 1e-06, "loss": 0.4255, "mean_token_accuracy": 0.8666738271713257, "num_tokens": 867807107.0, "step": 22744 }, { "epoch": 2.8933977865411524, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.578022956848145, "learning_rate": 1e-06, "loss": 0.4317, "mean_token_accuracy": 0.865475058555603, "num_tokens": 867850225.0, "step": 22745 }, { "epoch": 2.893524996819743, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.609275817871094, "learning_rate": 1e-06, "loss": 0.4107, "mean_token_accuracy": 0.8704102039337158, "num_tokens": 867890457.0, "step": 22746 }, { "epoch": 2.8936522070983335, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.616621017456055, "learning_rate": 1e-06, "loss": 0.3756, "mean_token_accuracy": 0.8801205158233643, "num_tokens": 867925880.0, "step": 22747 }, { "epoch": 2.893779417376924, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.556879997253418, "learning_rate": 1e-06, "loss": 0.356, "mean_token_accuracy": 0.8868306875228882, "num_tokens": 867963923.0, "step": 22748 }, { "epoch": 2.8939066276555145, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.557308197021484, "learning_rate": 1e-06, "loss": 0.3863, "mean_token_accuracy": 0.8774030208587646, "num_tokens": 868000489.0, "step": 22749 }, { "epoch": 2.894033837934105, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.659363746643066, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.876876950263977, "num_tokens": 868034001.0, "step": 22750 }, { "epoch": 2.8941610482126956, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.555861473083496, "learning_rate": 1e-06, "loss": 0.3955, "mean_token_accuracy": 0.8780397176742554, "num_tokens": 868074956.0, "step": 22751 }, { "epoch": 2.894288258491286, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.671602249145508, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8816468119621277, "num_tokens": 868110727.0, "step": 22752 }, { "epoch": 2.8944154687698767, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.567241668701172, "learning_rate": 1e-06, "loss": 0.4584, "mean_token_accuracy": 0.8559210896492004, "num_tokens": 868151480.0, "step": 22753 }, { "epoch": 2.894542679048467, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.604469299316406, "learning_rate": 1e-06, "loss": 0.4242, "mean_token_accuracy": 0.864345133304596, "num_tokens": 868190620.0, "step": 22754 }, { "epoch": 2.8946698893270577, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.58922004699707, "learning_rate": 1e-06, "loss": 0.4054, "mean_token_accuracy": 0.8714357614517212, "num_tokens": 868230598.0, "step": 22755 }, { "epoch": 2.8947970996056482, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.597854614257812, "learning_rate": 1e-06, "loss": 0.4324, "mean_token_accuracy": 0.8683228492736816, "num_tokens": 868265933.0, "step": 22756 }, { "epoch": 2.8949243098842388, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.59579849243164, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8761788606643677, "num_tokens": 868300282.0, "step": 22757 }, { "epoch": 2.8950515201628293, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.665871620178223, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8884469270706177, "num_tokens": 868331031.0, "step": 22758 }, { "epoch": 2.89517873044142, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.604368209838867, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8767857551574707, "num_tokens": 868368268.0, "step": 22759 }, { "epoch": 2.89530594072001, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.577473640441895, "learning_rate": 1e-06, "loss": 0.447, "mean_token_accuracy": 0.8594734072685242, "num_tokens": 868407551.0, "step": 22760 }, { "epoch": 2.895433150998601, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.613486289978027, "learning_rate": 1e-06, "loss": 0.4057, "mean_token_accuracy": 0.8713834881782532, "num_tokens": 868449301.0, "step": 22761 }, { "epoch": 2.895560361277191, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.550697326660156, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8713929653167725, "num_tokens": 868482853.0, "step": 22762 }, { "epoch": 2.895687571555782, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.61251449584961, "learning_rate": 1e-06, "loss": 0.4096, "mean_token_accuracy": 0.8691280484199524, "num_tokens": 868523960.0, "step": 22763 }, { "epoch": 2.895814781834372, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.617145538330078, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.866316020488739, "num_tokens": 868566447.0, "step": 22764 }, { "epoch": 2.895941992112963, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.509827613830566, "learning_rate": 1e-06, "loss": 0.3713, "mean_token_accuracy": 0.8835835456848145, "num_tokens": 868608342.0, "step": 22765 }, { "epoch": 2.896069202391553, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.574917793273926, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8853945136070251, "num_tokens": 868647354.0, "step": 22766 }, { "epoch": 2.8961964126701436, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.627872467041016, "learning_rate": 1e-06, "loss": 0.4606, "mean_token_accuracy": 0.8550463318824768, "num_tokens": 868690716.0, "step": 22767 }, { "epoch": 2.896323622948734, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.60957145690918, "learning_rate": 1e-06, "loss": 0.3788, "mean_token_accuracy": 0.8800475001335144, "num_tokens": 868728318.0, "step": 22768 }, { "epoch": 2.8964508332273247, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.583074569702148, "learning_rate": 1e-06, "loss": 0.3514, "mean_token_accuracy": 0.8900136947631836, "num_tokens": 868764323.0, "step": 22769 }, { "epoch": 2.896578043505915, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.668594360351562, "learning_rate": 1e-06, "loss": 0.4609, "mean_token_accuracy": 0.8569468259811401, "num_tokens": 868801371.0, "step": 22770 }, { "epoch": 2.8967052537845057, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.574271202087402, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8785348534584045, "num_tokens": 868846679.0, "step": 22771 }, { "epoch": 2.8968324640630962, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.64449691772461, "learning_rate": 1e-06, "loss": 0.4455, "mean_token_accuracy": 0.8613500595092773, "num_tokens": 868880076.0, "step": 22772 }, { "epoch": 2.8969596743416868, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.705352783203125, "learning_rate": 1e-06, "loss": 0.4056, "mean_token_accuracy": 0.8748207092285156, "num_tokens": 868917086.0, "step": 22773 }, { "epoch": 2.8970868846202773, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.561178207397461, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8800185918807983, "num_tokens": 868954952.0, "step": 22774 }, { "epoch": 2.897214094898868, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.674166679382324, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.8699278831481934, "num_tokens": 868990911.0, "step": 22775 }, { "epoch": 2.8973413051774584, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.602728843688965, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8843494057655334, "num_tokens": 869033398.0, "step": 22776 }, { "epoch": 2.897468515456049, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.586432456970215, "learning_rate": 1e-06, "loss": 0.4367, "mean_token_accuracy": 0.8609183430671692, "num_tokens": 869070513.0, "step": 22777 }, { "epoch": 2.8975957257346394, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.55351448059082, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8818320631980896, "num_tokens": 869113494.0, "step": 22778 }, { "epoch": 2.89772293601323, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.644257545471191, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.883409321308136, "num_tokens": 869147433.0, "step": 22779 }, { "epoch": 2.8978501462918205, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.583113670349121, "learning_rate": 1e-06, "loss": 0.4157, "mean_token_accuracy": 0.870150089263916, "num_tokens": 869186763.0, "step": 22780 }, { "epoch": 2.897977356570411, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.617177963256836, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8867030143737793, "num_tokens": 869222624.0, "step": 22781 }, { "epoch": 2.8981045668490015, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.585115432739258, "learning_rate": 1e-06, "loss": 0.3405, "mean_token_accuracy": 0.8943488597869873, "num_tokens": 869263027.0, "step": 22782 }, { "epoch": 2.898231777127592, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 14.587186813354492, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8691467046737671, "num_tokens": 869304232.0, "step": 22783 }, { "epoch": 2.8983589874061826, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.583728790283203, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8790345191955566, "num_tokens": 869340141.0, "step": 22784 }, { "epoch": 2.8984861976847727, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.614023208618164, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8655945062637329, "num_tokens": 869382428.0, "step": 22785 }, { "epoch": 2.8986134079633636, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.649869918823242, "learning_rate": 1e-06, "loss": 0.3766, "mean_token_accuracy": 0.8814497590065002, "num_tokens": 869418890.0, "step": 22786 }, { "epoch": 2.8987406182419537, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.57721996307373, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8758733868598938, "num_tokens": 869453704.0, "step": 22787 }, { "epoch": 2.8988678285205447, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.560282707214355, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8716838359832764, "num_tokens": 869491845.0, "step": 22788 }, { "epoch": 2.898995038799135, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.583101272583008, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8852615356445312, "num_tokens": 869523393.0, "step": 22789 }, { "epoch": 2.8991222490777258, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.627378463745117, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8716883659362793, "num_tokens": 869561927.0, "step": 22790 }, { "epoch": 2.899249459356316, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.552971839904785, "learning_rate": 1e-06, "loss": 0.3954, "mean_token_accuracy": 0.8733904361724854, "num_tokens": 869599568.0, "step": 22791 }, { "epoch": 2.8993766696349064, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.634622573852539, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8776516914367676, "num_tokens": 869636329.0, "step": 22792 }, { "epoch": 2.899503879913497, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.63581371307373, "learning_rate": 1e-06, "loss": 0.4135, "mean_token_accuracy": 0.8693420886993408, "num_tokens": 869679720.0, "step": 22793 }, { "epoch": 2.8996310901920874, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.614295959472656, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8770728707313538, "num_tokens": 869719527.0, "step": 22794 }, { "epoch": 2.899758300470678, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.61760425567627, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8789070844650269, "num_tokens": 869755139.0, "step": 22795 }, { "epoch": 2.8998855107492685, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.606588363647461, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8692255616188049, "num_tokens": 869791860.0, "step": 22796 }, { "epoch": 2.900012721027859, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.637555122375488, "learning_rate": 1e-06, "loss": 0.3382, "mean_token_accuracy": 0.8958828449249268, "num_tokens": 869827294.0, "step": 22797 }, { "epoch": 2.9001399313064495, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.628539085388184, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.869370698928833, "num_tokens": 869859398.0, "step": 22798 }, { "epoch": 2.90026714158504, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.625247955322266, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8739540576934814, "num_tokens": 869898215.0, "step": 22799 }, { "epoch": 2.9003943518636306, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.589350700378418, "learning_rate": 1e-06, "loss": 0.4266, "mean_token_accuracy": 0.8640971183776855, "num_tokens": 869939787.0, "step": 22800 }, { "epoch": 2.900521562142221, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.611701011657715, "learning_rate": 1e-06, "loss": 0.4345, "mean_token_accuracy": 0.865543007850647, "num_tokens": 869976790.0, "step": 22801 }, { "epoch": 2.9006487724208116, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.606405258178711, "learning_rate": 1e-06, "loss": 0.3825, "mean_token_accuracy": 0.8783769011497498, "num_tokens": 870013547.0, "step": 22802 }, { "epoch": 2.900775982699402, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.527685165405273, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.885353684425354, "num_tokens": 870049540.0, "step": 22803 }, { "epoch": 2.9009031929779927, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.590259552001953, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8807291388511658, "num_tokens": 870088247.0, "step": 22804 }, { "epoch": 2.9010304032565832, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.643150329589844, "learning_rate": 1e-06, "loss": 0.4048, "mean_token_accuracy": 0.8726232051849365, "num_tokens": 870129112.0, "step": 22805 }, { "epoch": 2.9011576135351738, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.624333381652832, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8740368485450745, "num_tokens": 870162082.0, "step": 22806 }, { "epoch": 2.9012848238137643, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.556413650512695, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8849591016769409, "num_tokens": 870199868.0, "step": 22807 }, { "epoch": 2.901412034092355, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.598671913146973, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.8790006637573242, "num_tokens": 870238121.0, "step": 22808 }, { "epoch": 2.9015392443709453, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.671765327453613, "learning_rate": 1e-06, "loss": 0.4271, "mean_token_accuracy": 0.8660305142402649, "num_tokens": 870278008.0, "step": 22809 }, { "epoch": 2.9016664546495354, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.541744232177734, "learning_rate": 1e-06, "loss": 0.3994, "mean_token_accuracy": 0.8717856407165527, "num_tokens": 870317671.0, "step": 22810 }, { "epoch": 2.9017936649281264, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.590193748474121, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8814432621002197, "num_tokens": 870352893.0, "step": 22811 }, { "epoch": 2.9019208752067165, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.607755661010742, "learning_rate": 1e-06, "loss": 0.3774, "mean_token_accuracy": 0.8828301429748535, "num_tokens": 870387513.0, "step": 22812 }, { "epoch": 2.9020480854853075, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.611627578735352, "learning_rate": 1e-06, "loss": 0.3762, "mean_token_accuracy": 0.8807631731033325, "num_tokens": 870428649.0, "step": 22813 }, { "epoch": 2.9021752957638975, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.574590682983398, "learning_rate": 1e-06, "loss": 0.3591, "mean_token_accuracy": 0.8883670568466187, "num_tokens": 870468329.0, "step": 22814 }, { "epoch": 2.902302506042488, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.580953598022461, "learning_rate": 1e-06, "loss": 0.4087, "mean_token_accuracy": 0.8735170960426331, "num_tokens": 870509273.0, "step": 22815 }, { "epoch": 2.9024297163210786, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.612364768981934, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8688730597496033, "num_tokens": 870544206.0, "step": 22816 }, { "epoch": 2.902556926599669, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.579177856445312, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8915422558784485, "num_tokens": 870578708.0, "step": 22817 }, { "epoch": 2.9026841368782597, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.568807601928711, "learning_rate": 1e-06, "loss": 0.3754, "mean_token_accuracy": 0.8813983798027039, "num_tokens": 870620599.0, "step": 22818 }, { "epoch": 2.90281134715685, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.604113578796387, "learning_rate": 1e-06, "loss": 0.3497, "mean_token_accuracy": 0.8886270523071289, "num_tokens": 870657828.0, "step": 22819 }, { "epoch": 2.9029385574354407, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.570438385009766, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8684711456298828, "num_tokens": 870694786.0, "step": 22820 }, { "epoch": 2.9030657677140312, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.558425903320312, "learning_rate": 1e-06, "loss": 0.3519, "mean_token_accuracy": 0.8915740251541138, "num_tokens": 870733850.0, "step": 22821 }, { "epoch": 2.9031929779926218, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.692761421203613, "learning_rate": 1e-06, "loss": 0.3992, "mean_token_accuracy": 0.8750582933425903, "num_tokens": 870771556.0, "step": 22822 }, { "epoch": 2.9033201882712123, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.572057723999023, "learning_rate": 1e-06, "loss": 0.3616, "mean_token_accuracy": 0.8857936859130859, "num_tokens": 870809335.0, "step": 22823 }, { "epoch": 2.903447398549803, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.61142635345459, "learning_rate": 1e-06, "loss": 0.4041, "mean_token_accuracy": 0.8702619075775146, "num_tokens": 870844239.0, "step": 22824 }, { "epoch": 2.9035746088283934, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.685358047485352, "learning_rate": 1e-06, "loss": 0.4018, "mean_token_accuracy": 0.8734430074691772, "num_tokens": 870879540.0, "step": 22825 }, { "epoch": 2.903701819106984, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.640273094177246, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8744344711303711, "num_tokens": 870917219.0, "step": 22826 }, { "epoch": 2.9038290293855744, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.56972885131836, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.875861644744873, "num_tokens": 870952332.0, "step": 22827 }, { "epoch": 2.903956239664165, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.667522430419922, "learning_rate": 1e-06, "loss": 0.4401, "mean_token_accuracy": 0.8633847832679749, "num_tokens": 870991609.0, "step": 22828 }, { "epoch": 2.9040834499427555, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.589203834533691, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8783231377601624, "num_tokens": 871027456.0, "step": 22829 }, { "epoch": 2.904210660221346, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.658963203430176, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8659446239471436, "num_tokens": 871071833.0, "step": 22830 }, { "epoch": 2.9043378704999365, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.592262268066406, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8730789422988892, "num_tokens": 871105446.0, "step": 22831 }, { "epoch": 2.904465080778527, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.648539543151855, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8730100393295288, "num_tokens": 871141408.0, "step": 22832 }, { "epoch": 2.9045922910571176, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.583312034606934, "learning_rate": 1e-06, "loss": 0.3887, "mean_token_accuracy": 0.8777439594268799, "num_tokens": 871173296.0, "step": 22833 }, { "epoch": 2.904719501335708, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.670710563659668, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8712363243103027, "num_tokens": 871210711.0, "step": 22834 }, { "epoch": 2.904846711614298, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.60667610168457, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.871569037437439, "num_tokens": 871251179.0, "step": 22835 }, { "epoch": 2.904973921892889, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.552950859069824, "learning_rate": 1e-06, "loss": 0.34, "mean_token_accuracy": 0.8937253952026367, "num_tokens": 871295758.0, "step": 22836 }, { "epoch": 2.9051011321714793, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.65130615234375, "learning_rate": 1e-06, "loss": 0.444, "mean_token_accuracy": 0.8594789505004883, "num_tokens": 871331145.0, "step": 22837 }, { "epoch": 2.9052283424500702, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.580406188964844, "learning_rate": 1e-06, "loss": 0.4698, "mean_token_accuracy": 0.8504521250724792, "num_tokens": 871372819.0, "step": 22838 }, { "epoch": 2.9053555527286603, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.546613693237305, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.8695093989372253, "num_tokens": 871414223.0, "step": 22839 }, { "epoch": 2.905482763007251, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.593526840209961, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8799021244049072, "num_tokens": 871455013.0, "step": 22840 }, { "epoch": 2.9056099732858414, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.584890365600586, "learning_rate": 1e-06, "loss": 0.4088, "mean_token_accuracy": 0.873429000377655, "num_tokens": 871495462.0, "step": 22841 }, { "epoch": 2.905737183564432, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.587446212768555, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8778868913650513, "num_tokens": 871534999.0, "step": 22842 }, { "epoch": 2.9058643938430224, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.57502555847168, "learning_rate": 1e-06, "loss": 0.395, "mean_token_accuracy": 0.8753988742828369, "num_tokens": 871569816.0, "step": 22843 }, { "epoch": 2.905991604121613, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.595483779907227, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8774375319480896, "num_tokens": 871610441.0, "step": 22844 }, { "epoch": 2.9061188144002035, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.54831600189209, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8607925772666931, "num_tokens": 871649005.0, "step": 22845 }, { "epoch": 2.906246024678794, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.674958229064941, "learning_rate": 1e-06, "loss": 0.4116, "mean_token_accuracy": 0.8702185153961182, "num_tokens": 871689263.0, "step": 22846 }, { "epoch": 2.9063732349573845, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.617012023925781, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.883647084236145, "num_tokens": 871725440.0, "step": 22847 }, { "epoch": 2.906500445235975, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.640113830566406, "learning_rate": 1e-06, "loss": 0.4122, "mean_token_accuracy": 0.8710861206054688, "num_tokens": 871767839.0, "step": 22848 }, { "epoch": 2.9066276555145656, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.592830657958984, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8790729641914368, "num_tokens": 871801516.0, "step": 22849 }, { "epoch": 2.906754865793156, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.645106315612793, "learning_rate": 1e-06, "loss": 0.4137, "mean_token_accuracy": 0.8699817061424255, "num_tokens": 871840359.0, "step": 22850 }, { "epoch": 2.9068820760717466, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.58003044128418, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8794495463371277, "num_tokens": 871879396.0, "step": 22851 }, { "epoch": 2.907009286350337, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.584639549255371, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8673133254051208, "num_tokens": 871915043.0, "step": 22852 }, { "epoch": 2.9071364966289277, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.582722663879395, "learning_rate": 1e-06, "loss": 0.4139, "mean_token_accuracy": 0.8699711561203003, "num_tokens": 871953391.0, "step": 22853 }, { "epoch": 2.9072637069075182, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.598546028137207, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8779929280281067, "num_tokens": 871989879.0, "step": 22854 }, { "epoch": 2.9073909171861088, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.631433486938477, "learning_rate": 1e-06, "loss": 0.3773, "mean_token_accuracy": 0.877530574798584, "num_tokens": 872022428.0, "step": 22855 }, { "epoch": 2.9075181274646993, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.591107368469238, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8789055943489075, "num_tokens": 872060644.0, "step": 22856 }, { "epoch": 2.90764533774329, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.616161346435547, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8711338043212891, "num_tokens": 872098230.0, "step": 22857 }, { "epoch": 2.90777254802188, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.659463882446289, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8634420037269592, "num_tokens": 872131217.0, "step": 22858 }, { "epoch": 2.907899758300471, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.675786018371582, "learning_rate": 1e-06, "loss": 0.3715, "mean_token_accuracy": 0.8836185932159424, "num_tokens": 872163912.0, "step": 22859 }, { "epoch": 2.908026968579061, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.598448753356934, "learning_rate": 1e-06, "loss": 0.4766, "mean_token_accuracy": 0.8503581285476685, "num_tokens": 872200694.0, "step": 22860 }, { "epoch": 2.908154178857652, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.649704933166504, "learning_rate": 1e-06, "loss": 0.3907, "mean_token_accuracy": 0.8745863437652588, "num_tokens": 872241156.0, "step": 22861 }, { "epoch": 2.908281389136242, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.545889854431152, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8774975538253784, "num_tokens": 872273048.0, "step": 22862 }, { "epoch": 2.908408599414833, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.614477157592773, "learning_rate": 1e-06, "loss": 0.3878, "mean_token_accuracy": 0.8781670928001404, "num_tokens": 872312165.0, "step": 22863 }, { "epoch": 2.908535809693423, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.624268531799316, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8771042823791504, "num_tokens": 872352740.0, "step": 22864 }, { "epoch": 2.9086630199720136, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.637282371520996, "learning_rate": 1e-06, "loss": 0.4307, "mean_token_accuracy": 0.8612822890281677, "num_tokens": 872397800.0, "step": 22865 }, { "epoch": 2.908790230250604, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.56673526763916, "learning_rate": 1e-06, "loss": 0.3683, "mean_token_accuracy": 0.8863989114761353, "num_tokens": 872429307.0, "step": 22866 }, { "epoch": 2.9089174405291947, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.604787826538086, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8721015453338623, "num_tokens": 872465897.0, "step": 22867 }, { "epoch": 2.909044650807785, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.621081352233887, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8655701875686646, "num_tokens": 872503252.0, "step": 22868 }, { "epoch": 2.9091718610863757, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.586996078491211, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8767791986465454, "num_tokens": 872546276.0, "step": 22869 }, { "epoch": 2.9092990713649662, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.622928619384766, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8726568818092346, "num_tokens": 872585690.0, "step": 22870 }, { "epoch": 2.9094262816435568, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.604192733764648, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8644417524337769, "num_tokens": 872621718.0, "step": 22871 }, { "epoch": 2.9095534919221473, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.643272399902344, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8790495991706848, "num_tokens": 872659349.0, "step": 22872 }, { "epoch": 2.909680702200738, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.644413948059082, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8700933456420898, "num_tokens": 872704468.0, "step": 22873 }, { "epoch": 2.9098079124793284, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.600153923034668, "learning_rate": 1e-06, "loss": 0.405, "mean_token_accuracy": 0.8750351667404175, "num_tokens": 872740003.0, "step": 22874 }, { "epoch": 2.909935122757919, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.635860443115234, "learning_rate": 1e-06, "loss": 0.4693, "mean_token_accuracy": 0.853486180305481, "num_tokens": 872780362.0, "step": 22875 }, { "epoch": 2.9100623330365094, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.680000305175781, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8827813863754272, "num_tokens": 872823719.0, "step": 22876 }, { "epoch": 2.9101895433151, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.632206916809082, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8788015842437744, "num_tokens": 872856276.0, "step": 22877 }, { "epoch": 2.9103167535936905, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.628758430480957, "learning_rate": 1e-06, "loss": 0.4166, "mean_token_accuracy": 0.867943286895752, "num_tokens": 872893688.0, "step": 22878 }, { "epoch": 2.910443963872281, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.609139442443848, "learning_rate": 1e-06, "loss": 0.3417, "mean_token_accuracy": 0.892578125, "num_tokens": 872931169.0, "step": 22879 }, { "epoch": 2.9105711741508715, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.624053001403809, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8692017793655396, "num_tokens": 872970836.0, "step": 22880 }, { "epoch": 2.910698384429462, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.59973430633545, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.8877334594726562, "num_tokens": 873010879.0, "step": 22881 }, { "epoch": 2.9108255947080526, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.616159439086914, "learning_rate": 1e-06, "loss": 0.396, "mean_token_accuracy": 0.8760324120521545, "num_tokens": 873058018.0, "step": 22882 }, { "epoch": 2.9109528049866427, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.642863273620605, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.86358243227005, "num_tokens": 873096259.0, "step": 22883 }, { "epoch": 2.9110800152652336, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.633228302001953, "learning_rate": 1e-06, "loss": 0.434, "mean_token_accuracy": 0.8646483421325684, "num_tokens": 873131955.0, "step": 22884 }, { "epoch": 2.9112072255438237, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.656983375549316, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8680769205093384, "num_tokens": 873171024.0, "step": 22885 }, { "epoch": 2.9113344358224147, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.612260818481445, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8787956833839417, "num_tokens": 873210568.0, "step": 22886 }, { "epoch": 2.9114616461010048, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.64564037322998, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8770947456359863, "num_tokens": 873250025.0, "step": 22887 }, { "epoch": 2.9115888563795957, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.543417930603027, "learning_rate": 1e-06, "loss": 0.429, "mean_token_accuracy": 0.866402268409729, "num_tokens": 873285284.0, "step": 22888 }, { "epoch": 2.911716066658186, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.68941593170166, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8769935965538025, "num_tokens": 873320102.0, "step": 22889 }, { "epoch": 2.9118432769367764, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.69782543182373, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8789359331130981, "num_tokens": 873354668.0, "step": 22890 }, { "epoch": 2.911970487215367, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.61662769317627, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8715980052947998, "num_tokens": 873387415.0, "step": 22891 }, { "epoch": 2.9120976974939574, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.593728065490723, "learning_rate": 1e-06, "loss": 0.3499, "mean_token_accuracy": 0.8911163210868835, "num_tokens": 873428679.0, "step": 22892 }, { "epoch": 2.912224907772548, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.680458068847656, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8754388689994812, "num_tokens": 873466973.0, "step": 22893 }, { "epoch": 2.9123521180511385, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.5905179977417, "learning_rate": 1e-06, "loss": 0.3494, "mean_token_accuracy": 0.8889569044113159, "num_tokens": 873510957.0, "step": 22894 }, { "epoch": 2.912479328329729, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.68244457244873, "learning_rate": 1e-06, "loss": 0.458, "mean_token_accuracy": 0.8552442193031311, "num_tokens": 873546456.0, "step": 22895 }, { "epoch": 2.9126065386083195, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.596570014953613, "learning_rate": 1e-06, "loss": 0.4406, "mean_token_accuracy": 0.8616011142730713, "num_tokens": 873591972.0, "step": 22896 }, { "epoch": 2.91273374888691, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.672365188598633, "learning_rate": 1e-06, "loss": 0.4019, "mean_token_accuracy": 0.8727900981903076, "num_tokens": 873629790.0, "step": 22897 }, { "epoch": 2.9128609591655006, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.63919448852539, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8739720582962036, "num_tokens": 873667419.0, "step": 22898 }, { "epoch": 2.912988169444091, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.628119468688965, "learning_rate": 1e-06, "loss": 0.3724, "mean_token_accuracy": 0.8842614889144897, "num_tokens": 873705834.0, "step": 22899 }, { "epoch": 2.9131153797226816, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.64742374420166, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8659255504608154, "num_tokens": 873748859.0, "step": 22900 }, { "epoch": 2.913242590001272, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.680047988891602, "learning_rate": 1e-06, "loss": 0.4245, "mean_token_accuracy": 0.8659977912902832, "num_tokens": 873788068.0, "step": 22901 }, { "epoch": 2.9133698002798627, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.64773941040039, "learning_rate": 1e-06, "loss": 0.373, "mean_token_accuracy": 0.8824989795684814, "num_tokens": 873827820.0, "step": 22902 }, { "epoch": 2.9134970105584532, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.681249618530273, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8704656958580017, "num_tokens": 873866271.0, "step": 22903 }, { "epoch": 2.9136242208370438, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.617741584777832, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8713136911392212, "num_tokens": 873907409.0, "step": 22904 }, { "epoch": 2.9137514311156343, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.736790657043457, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8802950382232666, "num_tokens": 873947940.0, "step": 22905 }, { "epoch": 2.913878641394225, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.62690544128418, "learning_rate": 1e-06, "loss": 0.4384, "mean_token_accuracy": 0.8599355816841125, "num_tokens": 873990062.0, "step": 22906 }, { "epoch": 2.9140058516728153, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.64467716217041, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8859169483184814, "num_tokens": 874026660.0, "step": 22907 }, { "epoch": 2.9141330619514054, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.62135124206543, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8808562755584717, "num_tokens": 874064935.0, "step": 22908 }, { "epoch": 2.9142602722299964, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.783024787902832, "learning_rate": 1e-06, "loss": 0.4066, "mean_token_accuracy": 0.8731657266616821, "num_tokens": 874102807.0, "step": 22909 }, { "epoch": 2.9143874825085865, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.600191116333008, "learning_rate": 1e-06, "loss": 0.3977, "mean_token_accuracy": 0.8761895298957825, "num_tokens": 874139688.0, "step": 22910 }, { "epoch": 2.9145146927871775, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.666954040527344, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.878851056098938, "num_tokens": 874184505.0, "step": 22911 }, { "epoch": 2.9146419030657675, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.606035232543945, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8739902973175049, "num_tokens": 874223210.0, "step": 22912 }, { "epoch": 2.914769113344358, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.693863868713379, "learning_rate": 1e-06, "loss": 0.4313, "mean_token_accuracy": 0.862069845199585, "num_tokens": 874263939.0, "step": 22913 }, { "epoch": 2.9148963236229486, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.591340065002441, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8818166255950928, "num_tokens": 874303946.0, "step": 22914 }, { "epoch": 2.915023533901539, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.617887496948242, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8764182329177856, "num_tokens": 874343701.0, "step": 22915 }, { "epoch": 2.9151507441801296, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.639636039733887, "learning_rate": 1e-06, "loss": 0.3942, "mean_token_accuracy": 0.8773110508918762, "num_tokens": 874380425.0, "step": 22916 }, { "epoch": 2.91527795445872, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.596892356872559, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8749157190322876, "num_tokens": 874417643.0, "step": 22917 }, { "epoch": 2.9154051647373107, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.651806831359863, "learning_rate": 1e-06, "loss": 0.4275, "mean_token_accuracy": 0.8680303692817688, "num_tokens": 874463939.0, "step": 22918 }, { "epoch": 2.9155323750159012, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.606254577636719, "learning_rate": 1e-06, "loss": 0.4119, "mean_token_accuracy": 0.8709884881973267, "num_tokens": 874503156.0, "step": 22919 }, { "epoch": 2.9156595852944918, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.690412521362305, "learning_rate": 1e-06, "loss": 0.4277, "mean_token_accuracy": 0.8644247055053711, "num_tokens": 874546754.0, "step": 22920 }, { "epoch": 2.9157867955730823, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.59243392944336, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8772959113121033, "num_tokens": 874590781.0, "step": 22921 }, { "epoch": 2.915914005851673, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.63303279876709, "learning_rate": 1e-06, "loss": 0.3667, "mean_token_accuracy": 0.8838144540786743, "num_tokens": 874629592.0, "step": 22922 }, { "epoch": 2.9160412161302633, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.5927095413208, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8816748261451721, "num_tokens": 874667372.0, "step": 22923 }, { "epoch": 2.916168426408854, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.60114860534668, "learning_rate": 1e-06, "loss": 0.4026, "mean_token_accuracy": 0.8718096613883972, "num_tokens": 874704687.0, "step": 22924 }, { "epoch": 2.9162956366874444, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.656574249267578, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8737189769744873, "num_tokens": 874740471.0, "step": 22925 }, { "epoch": 2.916422846966035, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.610361099243164, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8798953294754028, "num_tokens": 874780747.0, "step": 22926 }, { "epoch": 2.9165500572446255, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.644739151000977, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8787083625793457, "num_tokens": 874814204.0, "step": 22927 }, { "epoch": 2.916677267523216, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.712861061096191, "learning_rate": 1e-06, "loss": 0.4042, "mean_token_accuracy": 0.8738688826560974, "num_tokens": 874851872.0, "step": 22928 }, { "epoch": 2.9168044778018065, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.667195320129395, "learning_rate": 1e-06, "loss": 0.4332, "mean_token_accuracy": 0.8599351644515991, "num_tokens": 874884215.0, "step": 22929 }, { "epoch": 2.916931688080397, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.692585945129395, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8714298009872437, "num_tokens": 874919047.0, "step": 22930 }, { "epoch": 2.917058898358987, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.64952564239502, "learning_rate": 1e-06, "loss": 0.4103, "mean_token_accuracy": 0.8680206537246704, "num_tokens": 874956469.0, "step": 22931 }, { "epoch": 2.917186108637578, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.644546508789062, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8865394592285156, "num_tokens": 874994633.0, "step": 22932 }, { "epoch": 2.917313318916168, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.680290222167969, "learning_rate": 1e-06, "loss": 0.3642, "mean_token_accuracy": 0.8877083659172058, "num_tokens": 875028530.0, "step": 22933 }, { "epoch": 2.917440529194759, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.648494720458984, "learning_rate": 1e-06, "loss": 0.4209, "mean_token_accuracy": 0.867094874382019, "num_tokens": 875059785.0, "step": 22934 }, { "epoch": 2.9175677394733492, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.618696212768555, "learning_rate": 1e-06, "loss": 0.4027, "mean_token_accuracy": 0.8722536563873291, "num_tokens": 875096057.0, "step": 22935 }, { "epoch": 2.91769494975194, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.583921432495117, "learning_rate": 1e-06, "loss": 0.377, "mean_token_accuracy": 0.8804492354393005, "num_tokens": 875141988.0, "step": 22936 }, { "epoch": 2.9178221600305303, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.72664737701416, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8793772459030151, "num_tokens": 875176973.0, "step": 22937 }, { "epoch": 2.917949370309121, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.613663673400879, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8829180002212524, "num_tokens": 875209271.0, "step": 22938 }, { "epoch": 2.9180765805877114, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.67647647857666, "learning_rate": 1e-06, "loss": 0.3864, "mean_token_accuracy": 0.8793653249740601, "num_tokens": 875247460.0, "step": 22939 }, { "epoch": 2.918203790866302, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.585887908935547, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8723601698875427, "num_tokens": 875286013.0, "step": 22940 }, { "epoch": 2.9183310011448924, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.636577606201172, "learning_rate": 1e-06, "loss": 0.4165, "mean_token_accuracy": 0.8714866042137146, "num_tokens": 875324547.0, "step": 22941 }, { "epoch": 2.918458211423483, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.600556373596191, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8659260869026184, "num_tokens": 875366655.0, "step": 22942 }, { "epoch": 2.9185854217020735, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.623086929321289, "learning_rate": 1e-06, "loss": 0.4264, "mean_token_accuracy": 0.8683494329452515, "num_tokens": 875408571.0, "step": 22943 }, { "epoch": 2.918712631980664, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.579553604125977, "learning_rate": 1e-06, "loss": 0.4201, "mean_token_accuracy": 0.8687995672225952, "num_tokens": 875450730.0, "step": 22944 }, { "epoch": 2.9188398422592545, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.660961151123047, "learning_rate": 1e-06, "loss": 0.392, "mean_token_accuracy": 0.8779890537261963, "num_tokens": 875486850.0, "step": 22945 }, { "epoch": 2.918967052537845, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.662240982055664, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.868652880191803, "num_tokens": 875528653.0, "step": 22946 }, { "epoch": 2.9190942628164356, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.64465618133545, "learning_rate": 1e-06, "loss": 0.3704, "mean_token_accuracy": 0.8848700523376465, "num_tokens": 875564155.0, "step": 22947 }, { "epoch": 2.919221473095026, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.721349716186523, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8726973533630371, "num_tokens": 875598548.0, "step": 22948 }, { "epoch": 2.9193486833736166, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.590836524963379, "learning_rate": 1e-06, "loss": 0.354, "mean_token_accuracy": 0.8896452784538269, "num_tokens": 875634587.0, "step": 22949 }, { "epoch": 2.919475893652207, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.686180114746094, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8734291195869446, "num_tokens": 875671956.0, "step": 22950 }, { "epoch": 2.9196031039307977, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.637742042541504, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.876004695892334, "num_tokens": 875706327.0, "step": 22951 }, { "epoch": 2.9197303142093882, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.61846923828125, "learning_rate": 1e-06, "loss": 0.382, "mean_token_accuracy": 0.8809249997138977, "num_tokens": 875749320.0, "step": 22952 }, { "epoch": 2.9198575244879788, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.609753608703613, "learning_rate": 1e-06, "loss": 0.3769, "mean_token_accuracy": 0.8840559720993042, "num_tokens": 875783871.0, "step": 22953 }, { "epoch": 2.9199847347665693, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.620391845703125, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8739216327667236, "num_tokens": 875821716.0, "step": 22954 }, { "epoch": 2.92011194504516, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.657172203063965, "learning_rate": 1e-06, "loss": 0.3783, "mean_token_accuracy": 0.8786081671714783, "num_tokens": 875854339.0, "step": 22955 }, { "epoch": 2.92023915532375, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.624215126037598, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8757978677749634, "num_tokens": 875889456.0, "step": 22956 }, { "epoch": 2.920366365602341, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.634994506835938, "learning_rate": 1e-06, "loss": 0.4283, "mean_token_accuracy": 0.8662580251693726, "num_tokens": 875926659.0, "step": 22957 }, { "epoch": 2.920493575880931, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.619890213012695, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8815897107124329, "num_tokens": 875972766.0, "step": 22958 }, { "epoch": 2.920620786159522, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.588313102722168, "learning_rate": 1e-06, "loss": 0.4177, "mean_token_accuracy": 0.8706179261207581, "num_tokens": 876011824.0, "step": 22959 }, { "epoch": 2.920747996438112, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.670137405395508, "learning_rate": 1e-06, "loss": 0.4617, "mean_token_accuracy": 0.8575599789619446, "num_tokens": 876050353.0, "step": 22960 }, { "epoch": 2.920875206716703, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.619954109191895, "learning_rate": 1e-06, "loss": 0.442, "mean_token_accuracy": 0.8599637746810913, "num_tokens": 876089358.0, "step": 22961 }, { "epoch": 2.921002416995293, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.691930770874023, "learning_rate": 1e-06, "loss": 0.4397, "mean_token_accuracy": 0.8614689111709595, "num_tokens": 876126287.0, "step": 22962 }, { "epoch": 2.9211296272738836, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.597097396850586, "learning_rate": 1e-06, "loss": 0.3844, "mean_token_accuracy": 0.8835826516151428, "num_tokens": 876165157.0, "step": 22963 }, { "epoch": 2.921256837552474, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.70120906829834, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8761084079742432, "num_tokens": 876201391.0, "step": 22964 }, { "epoch": 2.9213840478310646, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.585590362548828, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.881282389163971, "num_tokens": 876243004.0, "step": 22965 }, { "epoch": 2.921511258109655, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.607661247253418, "learning_rate": 1e-06, "loss": 0.3874, "mean_token_accuracy": 0.8768874406814575, "num_tokens": 876279245.0, "step": 22966 }, { "epoch": 2.9216384683882457, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.662519454956055, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8738643527030945, "num_tokens": 876312145.0, "step": 22967 }, { "epoch": 2.9217656786668362, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.595968246459961, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8756008744239807, "num_tokens": 876351354.0, "step": 22968 }, { "epoch": 2.9218928889454268, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.694281578063965, "learning_rate": 1e-06, "loss": 0.4427, "mean_token_accuracy": 0.8639593124389648, "num_tokens": 876388621.0, "step": 22969 }, { "epoch": 2.9220200992240173, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.680719375610352, "learning_rate": 1e-06, "loss": 0.4185, "mean_token_accuracy": 0.8685581088066101, "num_tokens": 876418377.0, "step": 22970 }, { "epoch": 2.922147309502608, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.617012023925781, "learning_rate": 1e-06, "loss": 0.3951, "mean_token_accuracy": 0.8767225742340088, "num_tokens": 876453351.0, "step": 22971 }, { "epoch": 2.9222745197811983, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.667824745178223, "learning_rate": 1e-06, "loss": 0.3752, "mean_token_accuracy": 0.880804181098938, "num_tokens": 876484830.0, "step": 22972 }, { "epoch": 2.922401730059789, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.707722663879395, "learning_rate": 1e-06, "loss": 0.3677, "mean_token_accuracy": 0.8845996856689453, "num_tokens": 876524318.0, "step": 22973 }, { "epoch": 2.9225289403383794, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.622468948364258, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8718911409378052, "num_tokens": 876562421.0, "step": 22974 }, { "epoch": 2.92265615061697, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.643013000488281, "learning_rate": 1e-06, "loss": 0.4268, "mean_token_accuracy": 0.868215799331665, "num_tokens": 876605681.0, "step": 22975 }, { "epoch": 2.9227833608955605, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.635379791259766, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8651080131530762, "num_tokens": 876647371.0, "step": 22976 }, { "epoch": 2.922910571174151, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.602019309997559, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8803268671035767, "num_tokens": 876685563.0, "step": 22977 }, { "epoch": 2.9230377814527415, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.68508243560791, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8699932098388672, "num_tokens": 876718404.0, "step": 22978 }, { "epoch": 2.923164991731332, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.677749633789062, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8774707317352295, "num_tokens": 876751888.0, "step": 22979 }, { "epoch": 2.9232922020099226, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.62018871307373, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8793210983276367, "num_tokens": 876786813.0, "step": 22980 }, { "epoch": 2.9234194122885127, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.675653457641602, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8836148977279663, "num_tokens": 876824511.0, "step": 22981 }, { "epoch": 2.9235466225671036, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.654854774475098, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.877480149269104, "num_tokens": 876856409.0, "step": 22982 }, { "epoch": 2.9236738328456937, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.671538352966309, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8609865307807922, "num_tokens": 876889204.0, "step": 22983 }, { "epoch": 2.9238010431242847, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.655760765075684, "learning_rate": 1e-06, "loss": 0.3975, "mean_token_accuracy": 0.8760300874710083, "num_tokens": 876927185.0, "step": 22984 }, { "epoch": 2.9239282534028748, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.668710708618164, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8714867830276489, "num_tokens": 876965744.0, "step": 22985 }, { "epoch": 2.9240554636814657, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.649378776550293, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8803101181983948, "num_tokens": 876997107.0, "step": 22986 }, { "epoch": 2.924182673960056, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.652172088623047, "learning_rate": 1e-06, "loss": 0.4194, "mean_token_accuracy": 0.8651780486106873, "num_tokens": 877037894.0, "step": 22987 }, { "epoch": 2.9243098842386464, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.631745338439941, "learning_rate": 1e-06, "loss": 0.4481, "mean_token_accuracy": 0.8620429039001465, "num_tokens": 877078724.0, "step": 22988 }, { "epoch": 2.924437094517237, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.684130668640137, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.8751891851425171, "num_tokens": 877119685.0, "step": 22989 }, { "epoch": 2.9245643047958274, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.6448974609375, "learning_rate": 1e-06, "loss": 0.39, "mean_token_accuracy": 0.8776980042457581, "num_tokens": 877157055.0, "step": 22990 }, { "epoch": 2.924691515074418, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.62761402130127, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8802388310432434, "num_tokens": 877188569.0, "step": 22991 }, { "epoch": 2.9248187253530085, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.685495376586914, "learning_rate": 1e-06, "loss": 0.4228, "mean_token_accuracy": 0.867188572883606, "num_tokens": 877222745.0, "step": 22992 }, { "epoch": 2.924945935631599, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.634736061096191, "learning_rate": 1e-06, "loss": 0.4507, "mean_token_accuracy": 0.8588807582855225, "num_tokens": 877262510.0, "step": 22993 }, { "epoch": 2.9250731459101895, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.680750846862793, "learning_rate": 1e-06, "loss": 0.4198, "mean_token_accuracy": 0.8672648668289185, "num_tokens": 877300164.0, "step": 22994 }, { "epoch": 2.92520035618878, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.57758903503418, "learning_rate": 1e-06, "loss": 0.3732, "mean_token_accuracy": 0.8830982446670532, "num_tokens": 877337187.0, "step": 22995 }, { "epoch": 2.9253275664673706, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.600353240966797, "learning_rate": 1e-06, "loss": 0.4558, "mean_token_accuracy": 0.8557133674621582, "num_tokens": 877375902.0, "step": 22996 }, { "epoch": 2.925454776745961, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.698052406311035, "learning_rate": 1e-06, "loss": 0.4032, "mean_token_accuracy": 0.873124361038208, "num_tokens": 877414406.0, "step": 22997 }, { "epoch": 2.9255819870245516, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.661264419555664, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8803340196609497, "num_tokens": 877448977.0, "step": 22998 }, { "epoch": 2.925709197303142, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.608702659606934, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8814488649368286, "num_tokens": 877486766.0, "step": 22999 }, { "epoch": 2.9258364075817327, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.629351615905762, "learning_rate": 1e-06, "loss": 0.3602, "mean_token_accuracy": 0.8874189853668213, "num_tokens": 877519320.0, "step": 23000 }, { "epoch": 2.925963617860323, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.626275062561035, "learning_rate": 1e-06, "loss": 0.4774, "mean_token_accuracy": 0.8507289886474609, "num_tokens": 877557913.0, "step": 23001 }, { "epoch": 2.9260908281389137, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.674149513244629, "learning_rate": 1e-06, "loss": 0.3964, "mean_token_accuracy": 0.8743818998336792, "num_tokens": 877591197.0, "step": 23002 }, { "epoch": 2.9262180384175043, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.60515022277832, "learning_rate": 1e-06, "loss": 0.348, "mean_token_accuracy": 0.8904262781143188, "num_tokens": 877625018.0, "step": 23003 }, { "epoch": 2.926345248696095, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.56191349029541, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.879146158695221, "num_tokens": 877662433.0, "step": 23004 }, { "epoch": 2.9264724589746853, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.629261016845703, "learning_rate": 1e-06, "loss": 0.3749, "mean_token_accuracy": 0.8857301473617554, "num_tokens": 877703576.0, "step": 23005 }, { "epoch": 2.9265996692532754, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.645160675048828, "learning_rate": 1e-06, "loss": 0.4086, "mean_token_accuracy": 0.8705598711967468, "num_tokens": 877734025.0, "step": 23006 }, { "epoch": 2.9267268795318664, "ewc_loss": 0.039794921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 14.5654296875, "learning_rate": 1e-06, "loss": 0.3621, "mean_token_accuracy": 0.8837459683418274, "num_tokens": 877771942.0, "step": 23007 }, { "epoch": 2.9268540898104565, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.707974433898926, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8778237104415894, "num_tokens": 877815662.0, "step": 23008 }, { "epoch": 2.9269813000890474, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.66506576538086, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8596935272216797, "num_tokens": 877855398.0, "step": 23009 }, { "epoch": 2.9271085103676375, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.650382995605469, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8688145875930786, "num_tokens": 877898544.0, "step": 23010 }, { "epoch": 2.927235720646228, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.674125671386719, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8730682134628296, "num_tokens": 877937914.0, "step": 23011 }, { "epoch": 2.9273629309248186, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.610812187194824, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8576219081878662, "num_tokens": 877979266.0, "step": 23012 }, { "epoch": 2.927490141203409, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.671579360961914, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8788869976997375, "num_tokens": 878019762.0, "step": 23013 }, { "epoch": 2.9276173514819996, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.709349632263184, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8695075511932373, "num_tokens": 878062174.0, "step": 23014 }, { "epoch": 2.92774456176059, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.66840648651123, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8694313764572144, "num_tokens": 878098197.0, "step": 23015 }, { "epoch": 2.9278717720391807, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.66271686553955, "learning_rate": 1e-06, "loss": 0.3819, "mean_token_accuracy": 0.8798365592956543, "num_tokens": 878130478.0, "step": 23016 }, { "epoch": 2.9279989823177712, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.699793815612793, "learning_rate": 1e-06, "loss": 0.4504, "mean_token_accuracy": 0.857054591178894, "num_tokens": 878175670.0, "step": 23017 }, { "epoch": 2.9281261925963618, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.687155723571777, "learning_rate": 1e-06, "loss": 0.4326, "mean_token_accuracy": 0.8630655407905579, "num_tokens": 878216415.0, "step": 23018 }, { "epoch": 2.9282534028749523, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.681923866271973, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.8730465173721313, "num_tokens": 878249582.0, "step": 23019 }, { "epoch": 2.928380613153543, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.67097282409668, "learning_rate": 1e-06, "loss": 0.3631, "mean_token_accuracy": 0.8845980167388916, "num_tokens": 878290859.0, "step": 23020 }, { "epoch": 2.9285078234321333, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.646709442138672, "learning_rate": 1e-06, "loss": 0.3527, "mean_token_accuracy": 0.8904333710670471, "num_tokens": 878322463.0, "step": 23021 }, { "epoch": 2.928635033710724, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.68635368347168, "learning_rate": 1e-06, "loss": 0.3655, "mean_token_accuracy": 0.8839467763900757, "num_tokens": 878365754.0, "step": 23022 }, { "epoch": 2.9287622439893144, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.648810386657715, "learning_rate": 1e-06, "loss": 0.3613, "mean_token_accuracy": 0.8870729804039001, "num_tokens": 878410325.0, "step": 23023 }, { "epoch": 2.928889454267905, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.69835376739502, "learning_rate": 1e-06, "loss": 0.3791, "mean_token_accuracy": 0.8797197341918945, "num_tokens": 878451604.0, "step": 23024 }, { "epoch": 2.9290166645464955, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.643582344055176, "learning_rate": 1e-06, "loss": 0.3961, "mean_token_accuracy": 0.8775825500488281, "num_tokens": 878485174.0, "step": 23025 }, { "epoch": 2.929143874825086, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.751203536987305, "learning_rate": 1e-06, "loss": 0.4293, "mean_token_accuracy": 0.8666141033172607, "num_tokens": 878525867.0, "step": 23026 }, { "epoch": 2.9292710851036765, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.691889762878418, "learning_rate": 1e-06, "loss": 0.4395, "mean_token_accuracy": 0.8612338304519653, "num_tokens": 878561327.0, "step": 23027 }, { "epoch": 2.929398295382267, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.669363021850586, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8808650970458984, "num_tokens": 878603045.0, "step": 23028 }, { "epoch": 2.929525505660857, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.60069751739502, "learning_rate": 1e-06, "loss": 0.3768, "mean_token_accuracy": 0.8800152540206909, "num_tokens": 878645051.0, "step": 23029 }, { "epoch": 2.929652715939448, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.66073989868164, "learning_rate": 1e-06, "loss": 0.4599, "mean_token_accuracy": 0.8569735884666443, "num_tokens": 878682003.0, "step": 23030 }, { "epoch": 2.929779926218038, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.622336387634277, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8796991109848022, "num_tokens": 878720509.0, "step": 23031 }, { "epoch": 2.929907136496629, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.639451026916504, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8799026608467102, "num_tokens": 878757950.0, "step": 23032 }, { "epoch": 2.9300343467752192, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.561553955078125, "learning_rate": 1e-06, "loss": 0.4252, "mean_token_accuracy": 0.8674623966217041, "num_tokens": 878801477.0, "step": 23033 }, { "epoch": 2.93016155705381, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.69900894165039, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8670858144760132, "num_tokens": 878838802.0, "step": 23034 }, { "epoch": 2.9302887673324003, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.63820743560791, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.870294451713562, "num_tokens": 878875566.0, "step": 23035 }, { "epoch": 2.930415977610991, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.691665649414062, "learning_rate": 1e-06, "loss": 0.431, "mean_token_accuracy": 0.8672257661819458, "num_tokens": 878913463.0, "step": 23036 }, { "epoch": 2.9305431878895813, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.655326843261719, "learning_rate": 1e-06, "loss": 0.4355, "mean_token_accuracy": 0.8629026412963867, "num_tokens": 878950769.0, "step": 23037 }, { "epoch": 2.930670398168172, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.699110984802246, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8852034211158752, "num_tokens": 878989041.0, "step": 23038 }, { "epoch": 2.9307976084467624, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.672613143920898, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8820008635520935, "num_tokens": 879020195.0, "step": 23039 }, { "epoch": 2.930924818725353, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.581914901733398, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.8720545768737793, "num_tokens": 879057818.0, "step": 23040 }, { "epoch": 2.9310520290039435, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.690709114074707, "learning_rate": 1e-06, "loss": 0.4079, "mean_token_accuracy": 0.8713223338127136, "num_tokens": 879094747.0, "step": 23041 }, { "epoch": 2.931179239282534, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.682734489440918, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8652380704879761, "num_tokens": 879131127.0, "step": 23042 }, { "epoch": 2.9313064495611245, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.697615623474121, "learning_rate": 1e-06, "loss": 0.3896, "mean_token_accuracy": 0.8742917776107788, "num_tokens": 879168541.0, "step": 23043 }, { "epoch": 2.931433659839715, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.606928825378418, "learning_rate": 1e-06, "loss": 0.4156, "mean_token_accuracy": 0.8715164065361023, "num_tokens": 879199409.0, "step": 23044 }, { "epoch": 2.9315608701183056, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.66594123840332, "learning_rate": 1e-06, "loss": 0.3572, "mean_token_accuracy": 0.887471079826355, "num_tokens": 879236447.0, "step": 23045 }, { "epoch": 2.931688080396896, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.622594833374023, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8750096559524536, "num_tokens": 879275596.0, "step": 23046 }, { "epoch": 2.9318152906754866, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.662519454956055, "learning_rate": 1e-06, "loss": 0.4055, "mean_token_accuracy": 0.8732589483261108, "num_tokens": 879315620.0, "step": 23047 }, { "epoch": 2.931942500954077, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.592212677001953, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8766292333602905, "num_tokens": 879356220.0, "step": 23048 }, { "epoch": 2.9320697112326677, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.656254768371582, "learning_rate": 1e-06, "loss": 0.452, "mean_token_accuracy": 0.8599064350128174, "num_tokens": 879397544.0, "step": 23049 }, { "epoch": 2.932196921511258, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.633148193359375, "learning_rate": 1e-06, "loss": 0.3744, "mean_token_accuracy": 0.8843992948532104, "num_tokens": 879438165.0, "step": 23050 }, { "epoch": 2.9323241317898487, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.644075393676758, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8758490085601807, "num_tokens": 879473601.0, "step": 23051 }, { "epoch": 2.9324513420684393, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.645631790161133, "learning_rate": 1e-06, "loss": 0.4336, "mean_token_accuracy": 0.8664460182189941, "num_tokens": 879510998.0, "step": 23052 }, { "epoch": 2.93257855234703, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.711408615112305, "learning_rate": 1e-06, "loss": 0.3837, "mean_token_accuracy": 0.8793936967849731, "num_tokens": 879551469.0, "step": 23053 }, { "epoch": 2.93270576262562, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.660083770751953, "learning_rate": 1e-06, "loss": 0.4193, "mean_token_accuracy": 0.8675812482833862, "num_tokens": 879585620.0, "step": 23054 }, { "epoch": 2.932832972904211, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.636845588684082, "learning_rate": 1e-06, "loss": 0.3669, "mean_token_accuracy": 0.8839223384857178, "num_tokens": 879621749.0, "step": 23055 }, { "epoch": 2.932960183182801, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.658424377441406, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8777226805686951, "num_tokens": 879664529.0, "step": 23056 }, { "epoch": 2.933087393461392, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.678092002868652, "learning_rate": 1e-06, "loss": 0.3831, "mean_token_accuracy": 0.8793792724609375, "num_tokens": 879696293.0, "step": 23057 }, { "epoch": 2.933214603739982, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.644047737121582, "learning_rate": 1e-06, "loss": 0.3785, "mean_token_accuracy": 0.8790143728256226, "num_tokens": 879733562.0, "step": 23058 }, { "epoch": 2.933341814018573, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.705950736999512, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8715661764144897, "num_tokens": 879771740.0, "step": 23059 }, { "epoch": 2.933469024297163, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.685039520263672, "learning_rate": 1e-06, "loss": 0.4126, "mean_token_accuracy": 0.8701806664466858, "num_tokens": 879805578.0, "step": 23060 }, { "epoch": 2.9335962345757536, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.611713409423828, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8718691468238831, "num_tokens": 879845175.0, "step": 23061 }, { "epoch": 2.933723444854344, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.672175407409668, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8852404356002808, "num_tokens": 879878975.0, "step": 23062 }, { "epoch": 2.9338506551329346, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.673826217651367, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8659933805465698, "num_tokens": 879919373.0, "step": 23063 }, { "epoch": 2.933977865411525, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.719993591308594, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.8539764881134033, "num_tokens": 879953821.0, "step": 23064 }, { "epoch": 2.9341050756901157, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.641027450561523, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.8657135367393494, "num_tokens": 879993505.0, "step": 23065 }, { "epoch": 2.934232285968706, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.699904441833496, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8788619041442871, "num_tokens": 880030124.0, "step": 23066 }, { "epoch": 2.9343594962472968, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.71009349822998, "learning_rate": 1e-06, "loss": 0.465, "mean_token_accuracy": 0.8521769046783447, "num_tokens": 880059480.0, "step": 23067 }, { "epoch": 2.9344867065258873, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.645588874816895, "learning_rate": 1e-06, "loss": 0.3902, "mean_token_accuracy": 0.8752588033676147, "num_tokens": 880093402.0, "step": 23068 }, { "epoch": 2.934613916804478, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.66591739654541, "learning_rate": 1e-06, "loss": 0.3999, "mean_token_accuracy": 0.8725616931915283, "num_tokens": 880127874.0, "step": 23069 }, { "epoch": 2.9347411270830683, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.670016288757324, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8837132453918457, "num_tokens": 880165755.0, "step": 23070 }, { "epoch": 2.934868337361659, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.715041160583496, "learning_rate": 1e-06, "loss": 0.4256, "mean_token_accuracy": 0.8682520389556885, "num_tokens": 880203931.0, "step": 23071 }, { "epoch": 2.9349955476402494, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.625865936279297, "learning_rate": 1e-06, "loss": 0.391, "mean_token_accuracy": 0.8770653009414673, "num_tokens": 880242844.0, "step": 23072 }, { "epoch": 2.93512275791884, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.6914701461792, "learning_rate": 1e-06, "loss": 0.4416, "mean_token_accuracy": 0.8588738441467285, "num_tokens": 880280068.0, "step": 23073 }, { "epoch": 2.9352499681974304, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.671131134033203, "learning_rate": 1e-06, "loss": 0.3815, "mean_token_accuracy": 0.8798145651817322, "num_tokens": 880314044.0, "step": 23074 }, { "epoch": 2.935377178476021, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.667349815368652, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8709204196929932, "num_tokens": 880355077.0, "step": 23075 }, { "epoch": 2.9355043887546115, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.637587547302246, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8888161182403564, "num_tokens": 880388906.0, "step": 23076 }, { "epoch": 2.935631599033202, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.701757431030273, "learning_rate": 1e-06, "loss": 0.3718, "mean_token_accuracy": 0.8852452039718628, "num_tokens": 880426834.0, "step": 23077 }, { "epoch": 2.9357588093117926, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.647595405578613, "learning_rate": 1e-06, "loss": 0.4046, "mean_token_accuracy": 0.872267484664917, "num_tokens": 880463477.0, "step": 23078 }, { "epoch": 2.9358860195903826, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.769221305847168, "learning_rate": 1e-06, "loss": 0.4464, "mean_token_accuracy": 0.8578097224235535, "num_tokens": 880500884.0, "step": 23079 }, { "epoch": 2.9360132298689736, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.692163467407227, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8685047030448914, "num_tokens": 880539724.0, "step": 23080 }, { "epoch": 2.9361404401475637, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.72439193725586, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8669821619987488, "num_tokens": 880577820.0, "step": 23081 }, { "epoch": 2.9362676504261547, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.65652084350586, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8773194551467896, "num_tokens": 880612055.0, "step": 23082 }, { "epoch": 2.9363948607047448, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.69461441040039, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.881604790687561, "num_tokens": 880654907.0, "step": 23083 }, { "epoch": 2.9365220709833357, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.656883239746094, "learning_rate": 1e-06, "loss": 0.3448, "mean_token_accuracy": 0.8897958397865295, "num_tokens": 880692418.0, "step": 23084 }, { "epoch": 2.936649281261926, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.71408748626709, "learning_rate": 1e-06, "loss": 0.4371, "mean_token_accuracy": 0.8604873418807983, "num_tokens": 880727742.0, "step": 23085 }, { "epoch": 2.9367764915405163, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.695836067199707, "learning_rate": 1e-06, "loss": 0.3659, "mean_token_accuracy": 0.88579261302948, "num_tokens": 880766673.0, "step": 23086 }, { "epoch": 2.936903701819107, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.633066177368164, "learning_rate": 1e-06, "loss": 0.4257, "mean_token_accuracy": 0.8669031262397766, "num_tokens": 880810390.0, "step": 23087 }, { "epoch": 2.9370309120976974, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.74583911895752, "learning_rate": 1e-06, "loss": 0.4108, "mean_token_accuracy": 0.8694961071014404, "num_tokens": 880846920.0, "step": 23088 }, { "epoch": 2.937158122376288, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.678300857543945, "learning_rate": 1e-06, "loss": 0.3727, "mean_token_accuracy": 0.880778968334198, "num_tokens": 880885867.0, "step": 23089 }, { "epoch": 2.9372853326548785, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.669398307800293, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8850209712982178, "num_tokens": 880928408.0, "step": 23090 }, { "epoch": 2.937412542933469, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.685656547546387, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8738996982574463, "num_tokens": 880960681.0, "step": 23091 }, { "epoch": 2.9375397532120595, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.676302909851074, "learning_rate": 1e-06, "loss": 0.4281, "mean_token_accuracy": 0.8625900149345398, "num_tokens": 881000358.0, "step": 23092 }, { "epoch": 2.93766696349065, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.676491737365723, "learning_rate": 1e-06, "loss": 0.3957, "mean_token_accuracy": 0.8761483430862427, "num_tokens": 881037765.0, "step": 23093 }, { "epoch": 2.9377941737692406, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.629222869873047, "learning_rate": 1e-06, "loss": 0.3604, "mean_token_accuracy": 0.8881374597549438, "num_tokens": 881076091.0, "step": 23094 }, { "epoch": 2.937921384047831, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.671137809753418, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8734967112541199, "num_tokens": 881112784.0, "step": 23095 }, { "epoch": 2.9380485943264216, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.673723220825195, "learning_rate": 1e-06, "loss": 0.3841, "mean_token_accuracy": 0.8798006772994995, "num_tokens": 881149938.0, "step": 23096 }, { "epoch": 2.938175804605012, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.717365264892578, "learning_rate": 1e-06, "loss": 0.4239, "mean_token_accuracy": 0.8658019304275513, "num_tokens": 881187541.0, "step": 23097 }, { "epoch": 2.9383030148836027, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.678436279296875, "learning_rate": 1e-06, "loss": 0.3904, "mean_token_accuracy": 0.8765588998794556, "num_tokens": 881221920.0, "step": 23098 }, { "epoch": 2.938430225162193, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.683371543884277, "learning_rate": 1e-06, "loss": 0.4424, "mean_token_accuracy": 0.8593599796295166, "num_tokens": 881264804.0, "step": 23099 }, { "epoch": 2.9385574354407837, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.678057670593262, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.8663881421089172, "num_tokens": 881304482.0, "step": 23100 }, { "epoch": 2.9386846457193743, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.714428901672363, "learning_rate": 1e-06, "loss": 0.4352, "mean_token_accuracy": 0.8644112944602966, "num_tokens": 881340822.0, "step": 23101 }, { "epoch": 2.938811855997965, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.707231521606445, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8718385696411133, "num_tokens": 881374352.0, "step": 23102 }, { "epoch": 2.9389390662765553, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.720098495483398, "learning_rate": 1e-06, "loss": 0.4123, "mean_token_accuracy": 0.8691571950912476, "num_tokens": 881413176.0, "step": 23103 }, { "epoch": 2.9390662765551454, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.702604293823242, "learning_rate": 1e-06, "loss": 0.4154, "mean_token_accuracy": 0.8699132204055786, "num_tokens": 881454309.0, "step": 23104 }, { "epoch": 2.9391934868337364, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.70838451385498, "learning_rate": 1e-06, "loss": 0.3943, "mean_token_accuracy": 0.874619722366333, "num_tokens": 881490702.0, "step": 23105 }, { "epoch": 2.9393206971123265, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.650483131408691, "learning_rate": 1e-06, "loss": 0.4038, "mean_token_accuracy": 0.8801543712615967, "num_tokens": 881530394.0, "step": 23106 }, { "epoch": 2.9394479073909174, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.752397537231445, "learning_rate": 1e-06, "loss": 0.3547, "mean_token_accuracy": 0.8877450227737427, "num_tokens": 881567286.0, "step": 23107 }, { "epoch": 2.9395751176695075, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.718348503112793, "learning_rate": 1e-06, "loss": 0.4353, "mean_token_accuracy": 0.8689883947372437, "num_tokens": 881607017.0, "step": 23108 }, { "epoch": 2.939702327948098, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.7091064453125, "learning_rate": 1e-06, "loss": 0.3921, "mean_token_accuracy": 0.8747812509536743, "num_tokens": 881645793.0, "step": 23109 }, { "epoch": 2.9398295382266886, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.639692306518555, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8769766092300415, "num_tokens": 881683596.0, "step": 23110 }, { "epoch": 2.939956748505279, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.704924583435059, "learning_rate": 1e-06, "loss": 0.4516, "mean_token_accuracy": 0.8582509756088257, "num_tokens": 881717607.0, "step": 23111 }, { "epoch": 2.9400839587838696, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.74985122680664, "learning_rate": 1e-06, "loss": 0.3585, "mean_token_accuracy": 0.8858853578567505, "num_tokens": 881750884.0, "step": 23112 }, { "epoch": 2.94021116906246, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.70081901550293, "learning_rate": 1e-06, "loss": 0.4186, "mean_token_accuracy": 0.868718147277832, "num_tokens": 881789345.0, "step": 23113 }, { "epoch": 2.9403383793410507, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.729305267333984, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.870628833770752, "num_tokens": 881828459.0, "step": 23114 }, { "epoch": 2.940465589619641, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.734562873840332, "learning_rate": 1e-06, "loss": 0.425, "mean_token_accuracy": 0.868156909942627, "num_tokens": 881865294.0, "step": 23115 }, { "epoch": 2.9405927998982317, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.768068313598633, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8755348920822144, "num_tokens": 881904936.0, "step": 23116 }, { "epoch": 2.9407200101768223, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.71442699432373, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.88509601354599, "num_tokens": 881936300.0, "step": 23117 }, { "epoch": 2.940847220455413, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.722222328186035, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8858155012130737, "num_tokens": 881974092.0, "step": 23118 }, { "epoch": 2.9409744307340033, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.710668563842773, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8752549290657043, "num_tokens": 882019339.0, "step": 23119 }, { "epoch": 2.941101641012594, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.660768508911133, "learning_rate": 1e-06, "loss": 0.3895, "mean_token_accuracy": 0.8776093125343323, "num_tokens": 882052837.0, "step": 23120 }, { "epoch": 2.9412288512911844, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.745348930358887, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8792421817779541, "num_tokens": 882088449.0, "step": 23121 }, { "epoch": 2.941356061569775, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.664929389953613, "learning_rate": 1e-06, "loss": 0.4181, "mean_token_accuracy": 0.8656631708145142, "num_tokens": 882124844.0, "step": 23122 }, { "epoch": 2.9414832718483654, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.712143898010254, "learning_rate": 1e-06, "loss": 0.3846, "mean_token_accuracy": 0.8764364123344421, "num_tokens": 882162246.0, "step": 23123 }, { "epoch": 2.941610482126956, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.679306983947754, "learning_rate": 1e-06, "loss": 0.4795, "mean_token_accuracy": 0.8625782132148743, "num_tokens": 882202825.0, "step": 23124 }, { "epoch": 2.9417376924055465, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.68393611907959, "learning_rate": 1e-06, "loss": 0.4375, "mean_token_accuracy": 0.8592337369918823, "num_tokens": 882237951.0, "step": 23125 }, { "epoch": 2.941864902684137, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.628511428833008, "learning_rate": 1e-06, "loss": 0.4052, "mean_token_accuracy": 0.873319149017334, "num_tokens": 882279147.0, "step": 23126 }, { "epoch": 2.941992112962727, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.651073455810547, "learning_rate": 1e-06, "loss": 0.3377, "mean_token_accuracy": 0.8940809369087219, "num_tokens": 882322777.0, "step": 23127 }, { "epoch": 2.942119323241318, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.686156272888184, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8791869878768921, "num_tokens": 882360681.0, "step": 23128 }, { "epoch": 2.942246533519908, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.661991119384766, "learning_rate": 1e-06, "loss": 0.4568, "mean_token_accuracy": 0.8538950085639954, "num_tokens": 882403823.0, "step": 23129 }, { "epoch": 2.942373743798499, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.679518699645996, "learning_rate": 1e-06, "loss": 0.384, "mean_token_accuracy": 0.8823581337928772, "num_tokens": 882443723.0, "step": 23130 }, { "epoch": 2.9425009540770892, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.723983764648438, "learning_rate": 1e-06, "loss": 0.4517, "mean_token_accuracy": 0.8598604798316956, "num_tokens": 882480610.0, "step": 23131 }, { "epoch": 2.94262816435568, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.670553207397461, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8765921592712402, "num_tokens": 882518789.0, "step": 23132 }, { "epoch": 2.9427553746342703, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.647368431091309, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8608887195587158, "num_tokens": 882556598.0, "step": 23133 }, { "epoch": 2.942882584912861, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.702741622924805, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8697423934936523, "num_tokens": 882593788.0, "step": 23134 }, { "epoch": 2.9430097951914513, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.611909866333008, "learning_rate": 1e-06, "loss": 0.3823, "mean_token_accuracy": 0.8805803060531616, "num_tokens": 882630861.0, "step": 23135 }, { "epoch": 2.943137005470042, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.603414535522461, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8816162347793579, "num_tokens": 882670950.0, "step": 23136 }, { "epoch": 2.9432642157486324, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.669058799743652, "learning_rate": 1e-06, "loss": 0.4327, "mean_token_accuracy": 0.8644001483917236, "num_tokens": 882705680.0, "step": 23137 }, { "epoch": 2.943391426027223, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.683633804321289, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8826160430908203, "num_tokens": 882737589.0, "step": 23138 }, { "epoch": 2.9435186363058135, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.70154094696045, "learning_rate": 1e-06, "loss": 0.4474, "mean_token_accuracy": 0.8590250015258789, "num_tokens": 882774502.0, "step": 23139 }, { "epoch": 2.943645846584404, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.65570068359375, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.8640021085739136, "num_tokens": 882815184.0, "step": 23140 }, { "epoch": 2.9437730568629945, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.640143394470215, "learning_rate": 1e-06, "loss": 0.363, "mean_token_accuracy": 0.8864246606826782, "num_tokens": 882856793.0, "step": 23141 }, { "epoch": 2.943900267141585, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.648051261901855, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8840052485466003, "num_tokens": 882896128.0, "step": 23142 }, { "epoch": 2.9440274774201756, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.732616424560547, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8814537525177002, "num_tokens": 882931253.0, "step": 23143 }, { "epoch": 2.944154687698766, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.608031272888184, "learning_rate": 1e-06, "loss": 0.3506, "mean_token_accuracy": 0.8893353343009949, "num_tokens": 882975072.0, "step": 23144 }, { "epoch": 2.9442818979773566, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.718238830566406, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8859193921089172, "num_tokens": 883013576.0, "step": 23145 }, { "epoch": 2.944409108255947, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.677082061767578, "learning_rate": 1e-06, "loss": 0.3868, "mean_token_accuracy": 0.8768042325973511, "num_tokens": 883046052.0, "step": 23146 }, { "epoch": 2.9445363185345377, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.685628890991211, "learning_rate": 1e-06, "loss": 0.4596, "mean_token_accuracy": 0.8555870652198792, "num_tokens": 883080398.0, "step": 23147 }, { "epoch": 2.944663528813128, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.673901557922363, "learning_rate": 1e-06, "loss": 0.4237, "mean_token_accuracy": 0.86519455909729, "num_tokens": 883120321.0, "step": 23148 }, { "epoch": 2.9447907390917187, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.679463386535645, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8811086416244507, "num_tokens": 883161285.0, "step": 23149 }, { "epoch": 2.9449179493703093, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.703289031982422, "learning_rate": 1e-06, "loss": 0.4028, "mean_token_accuracy": 0.8723508715629578, "num_tokens": 883199857.0, "step": 23150 }, { "epoch": 2.9450451596489, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.695292472839355, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8808127641677856, "num_tokens": 883238325.0, "step": 23151 }, { "epoch": 2.94517236992749, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.671302795410156, "learning_rate": 1e-06, "loss": 0.4217, "mean_token_accuracy": 0.8655307292938232, "num_tokens": 883281814.0, "step": 23152 }, { "epoch": 2.945299580206081, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.760857582092285, "learning_rate": 1e-06, "loss": 0.3803, "mean_token_accuracy": 0.8789572715759277, "num_tokens": 883317385.0, "step": 23153 }, { "epoch": 2.945426790484671, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.69676685333252, "learning_rate": 1e-06, "loss": 0.3737, "mean_token_accuracy": 0.881165623664856, "num_tokens": 883353498.0, "step": 23154 }, { "epoch": 2.945554000763262, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.643156051635742, "learning_rate": 1e-06, "loss": 0.3811, "mean_token_accuracy": 0.8825207948684692, "num_tokens": 883391709.0, "step": 23155 }, { "epoch": 2.945681211041852, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.698838233947754, "learning_rate": 1e-06, "loss": 0.4496, "mean_token_accuracy": 0.8585565090179443, "num_tokens": 883426329.0, "step": 23156 }, { "epoch": 2.945808421320443, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.69587516784668, "learning_rate": 1e-06, "loss": 0.4034, "mean_token_accuracy": 0.871654212474823, "num_tokens": 883465797.0, "step": 23157 }, { "epoch": 2.945935631599033, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.69947338104248, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.8771421909332275, "num_tokens": 883501632.0, "step": 23158 }, { "epoch": 2.9460628418776236, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.676803588867188, "learning_rate": 1e-06, "loss": 0.3686, "mean_token_accuracy": 0.8849416971206665, "num_tokens": 883537622.0, "step": 23159 }, { "epoch": 2.946190052156214, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.700071334838867, "learning_rate": 1e-06, "loss": 0.3983, "mean_token_accuracy": 0.8753372430801392, "num_tokens": 883576458.0, "step": 23160 }, { "epoch": 2.9463172624348046, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.670870780944824, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8647934198379517, "num_tokens": 883620929.0, "step": 23161 }, { "epoch": 2.946444472713395, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.69290542602539, "learning_rate": 1e-06, "loss": 0.4792, "mean_token_accuracy": 0.8502166867256165, "num_tokens": 883652589.0, "step": 23162 }, { "epoch": 2.9465716829919857, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.65779972076416, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8740966320037842, "num_tokens": 883691844.0, "step": 23163 }, { "epoch": 2.946698893270576, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.72472858428955, "learning_rate": 1e-06, "loss": 0.4618, "mean_token_accuracy": 0.8549517393112183, "num_tokens": 883732838.0, "step": 23164 }, { "epoch": 2.9468261035491667, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.645133018493652, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8714715838432312, "num_tokens": 883771415.0, "step": 23165 }, { "epoch": 2.9469533138277573, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.717484474182129, "learning_rate": 1e-06, "loss": 0.4458, "mean_token_accuracy": 0.8596967458724976, "num_tokens": 883808807.0, "step": 23166 }, { "epoch": 2.947080524106348, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.698705673217773, "learning_rate": 1e-06, "loss": 0.3516, "mean_token_accuracy": 0.8880170583724976, "num_tokens": 883844300.0, "step": 23167 }, { "epoch": 2.9472077343849383, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.658609390258789, "learning_rate": 1e-06, "loss": 0.3566, "mean_token_accuracy": 0.8876087665557861, "num_tokens": 883879337.0, "step": 23168 }, { "epoch": 2.947334944663529, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.687421798706055, "learning_rate": 1e-06, "loss": 0.43, "mean_token_accuracy": 0.8635236024856567, "num_tokens": 883925810.0, "step": 23169 }, { "epoch": 2.9474621549421194, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.709844589233398, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8836409449577332, "num_tokens": 883964753.0, "step": 23170 }, { "epoch": 2.94758936522071, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.745017051696777, "learning_rate": 1e-06, "loss": 0.4314, "mean_token_accuracy": 0.8670436143875122, "num_tokens": 884005737.0, "step": 23171 }, { "epoch": 2.9477165754993004, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.709016799926758, "learning_rate": 1e-06, "loss": 0.4182, "mean_token_accuracy": 0.8748205900192261, "num_tokens": 884043414.0, "step": 23172 }, { "epoch": 2.947843785777891, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.72750473022461, "learning_rate": 1e-06, "loss": 0.3852, "mean_token_accuracy": 0.8788557052612305, "num_tokens": 884077399.0, "step": 23173 }, { "epoch": 2.9479709960564815, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.708488464355469, "learning_rate": 1e-06, "loss": 0.406, "mean_token_accuracy": 0.8727971315383911, "num_tokens": 884116539.0, "step": 23174 }, { "epoch": 2.948098206335072, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.668499946594238, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8800523281097412, "num_tokens": 884149756.0, "step": 23175 }, { "epoch": 2.9482254166136626, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.753201484680176, "learning_rate": 1e-06, "loss": 0.3507, "mean_token_accuracy": 0.8934302926063538, "num_tokens": 884181960.0, "step": 23176 }, { "epoch": 2.9483526268922526, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.659844398498535, "learning_rate": 1e-06, "loss": 0.4114, "mean_token_accuracy": 0.8705235719680786, "num_tokens": 884223562.0, "step": 23177 }, { "epoch": 2.9484798371708436, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.701213836669922, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8802695274353027, "num_tokens": 884264541.0, "step": 23178 }, { "epoch": 2.9486070474494337, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.781254768371582, "learning_rate": 1e-06, "loss": 0.4451, "mean_token_accuracy": 0.8598299026489258, "num_tokens": 884301171.0, "step": 23179 }, { "epoch": 2.9487342577280247, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.61879825592041, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.8735146522521973, "num_tokens": 884332481.0, "step": 23180 }, { "epoch": 2.9488614680066147, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.778947830200195, "learning_rate": 1e-06, "loss": 0.3383, "mean_token_accuracy": 0.894866943359375, "num_tokens": 884371616.0, "step": 23181 }, { "epoch": 2.9489886782852053, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.710679054260254, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8635942935943604, "num_tokens": 884411245.0, "step": 23182 }, { "epoch": 2.949115888563796, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.750984191894531, "learning_rate": 1e-06, "loss": 0.412, "mean_token_accuracy": 0.8721465468406677, "num_tokens": 884445712.0, "step": 23183 }, { "epoch": 2.9492430988423863, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.751993179321289, "learning_rate": 1e-06, "loss": 0.3862, "mean_token_accuracy": 0.8776205778121948, "num_tokens": 884484935.0, "step": 23184 }, { "epoch": 2.949370309120977, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.656102180480957, "learning_rate": 1e-06, "loss": 0.3937, "mean_token_accuracy": 0.8747662305831909, "num_tokens": 884518714.0, "step": 23185 }, { "epoch": 2.9494975193995674, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.722098350524902, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8660557866096497, "num_tokens": 884560430.0, "step": 23186 }, { "epoch": 2.949624729678158, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.7340726852417, "learning_rate": 1e-06, "loss": 0.4175, "mean_token_accuracy": 0.8660529851913452, "num_tokens": 884595937.0, "step": 23187 }, { "epoch": 2.9497519399567484, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.694337844848633, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8684394955635071, "num_tokens": 884633201.0, "step": 23188 }, { "epoch": 2.949879150235339, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.700033187866211, "learning_rate": 1e-06, "loss": 0.3925, "mean_token_accuracy": 0.8753871917724609, "num_tokens": 884672602.0, "step": 23189 }, { "epoch": 2.9500063605139295, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.707722663879395, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8785449266433716, "num_tokens": 884708378.0, "step": 23190 }, { "epoch": 2.95013357079252, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.656545639038086, "learning_rate": 1e-06, "loss": 0.4051, "mean_token_accuracy": 0.8722784519195557, "num_tokens": 884747527.0, "step": 23191 }, { "epoch": 2.9502607810711106, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.691899299621582, "learning_rate": 1e-06, "loss": 0.3418, "mean_token_accuracy": 0.892126202583313, "num_tokens": 884781738.0, "step": 23192 }, { "epoch": 2.950387991349701, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.710317611694336, "learning_rate": 1e-06, "loss": 0.4343, "mean_token_accuracy": 0.8634111285209656, "num_tokens": 884818884.0, "step": 23193 }, { "epoch": 2.9505152016282916, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.718709945678711, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8843618631362915, "num_tokens": 884851751.0, "step": 23194 }, { "epoch": 2.950642411906882, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.63757038116455, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.8781919479370117, "num_tokens": 884892935.0, "step": 23195 }, { "epoch": 2.9507696221854727, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.691205024719238, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8779963850975037, "num_tokens": 884935820.0, "step": 23196 }, { "epoch": 2.950896832464063, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.599283218383789, "learning_rate": 1e-06, "loss": 0.4117, "mean_token_accuracy": 0.8726803064346313, "num_tokens": 884976042.0, "step": 23197 }, { "epoch": 2.9510240427426537, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.655977249145508, "learning_rate": 1e-06, "loss": 0.3383, "mean_token_accuracy": 0.893073558807373, "num_tokens": 885013511.0, "step": 23198 }, { "epoch": 2.9511512530212443, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.648507118225098, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8619902729988098, "num_tokens": 885047291.0, "step": 23199 }, { "epoch": 2.951278463299835, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.699369430541992, "learning_rate": 1e-06, "loss": 0.4142, "mean_token_accuracy": 0.8708382844924927, "num_tokens": 885083116.0, "step": 23200 }, { "epoch": 2.9514056735784253, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.707693099975586, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.869213342666626, "num_tokens": 885123282.0, "step": 23201 }, { "epoch": 2.9515328838570154, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.626630783081055, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.8788579702377319, "num_tokens": 885166277.0, "step": 23202 }, { "epoch": 2.9516600941356064, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.70185375213623, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8850032091140747, "num_tokens": 885206414.0, "step": 23203 }, { "epoch": 2.9517873044141965, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.724467277526855, "learning_rate": 1e-06, "loss": 0.3855, "mean_token_accuracy": 0.8822311162948608, "num_tokens": 885240977.0, "step": 23204 }, { "epoch": 2.9519145146927874, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.691508293151855, "learning_rate": 1e-06, "loss": 0.35, "mean_token_accuracy": 0.8894449472427368, "num_tokens": 885283190.0, "step": 23205 }, { "epoch": 2.9520417249713775, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.66780948638916, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.8885715007781982, "num_tokens": 885321340.0, "step": 23206 }, { "epoch": 2.952168935249968, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.712862968444824, "learning_rate": 1e-06, "loss": 0.3734, "mean_token_accuracy": 0.8832097053527832, "num_tokens": 885357524.0, "step": 23207 }, { "epoch": 2.9522961455285586, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.741748809814453, "learning_rate": 1e-06, "loss": 0.4653, "mean_token_accuracy": 0.8509669899940491, "num_tokens": 885389696.0, "step": 23208 }, { "epoch": 2.952423355807149, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.636602401733398, "learning_rate": 1e-06, "loss": 0.3523, "mean_token_accuracy": 0.8889954090118408, "num_tokens": 885429574.0, "step": 23209 }, { "epoch": 2.9525505660857396, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.708141326904297, "learning_rate": 1e-06, "loss": 0.3861, "mean_token_accuracy": 0.8793757557868958, "num_tokens": 885459961.0, "step": 23210 }, { "epoch": 2.95267777636433, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.620810508728027, "learning_rate": 1e-06, "loss": 0.4212, "mean_token_accuracy": 0.8682964444160461, "num_tokens": 885506725.0, "step": 23211 }, { "epoch": 2.9528049866429207, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.66810417175293, "learning_rate": 1e-06, "loss": 0.3906, "mean_token_accuracy": 0.8796027898788452, "num_tokens": 885549409.0, "step": 23212 }, { "epoch": 2.952932196921511, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.652911186218262, "learning_rate": 1e-06, "loss": 0.3881, "mean_token_accuracy": 0.8798122406005859, "num_tokens": 885589272.0, "step": 23213 }, { "epoch": 2.9530594072001017, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.735621452331543, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8558782935142517, "num_tokens": 885625951.0, "step": 23214 }, { "epoch": 2.9531866174786923, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.664837837219238, "learning_rate": 1e-06, "loss": 0.3926, "mean_token_accuracy": 0.8772192001342773, "num_tokens": 885665844.0, "step": 23215 }, { "epoch": 2.953313827757283, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.700411796569824, "learning_rate": 1e-06, "loss": 0.3998, "mean_token_accuracy": 0.8738638758659363, "num_tokens": 885705616.0, "step": 23216 }, { "epoch": 2.9534410380358733, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.662214279174805, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8865752220153809, "num_tokens": 885752362.0, "step": 23217 }, { "epoch": 2.953568248314464, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.678415298461914, "learning_rate": 1e-06, "loss": 0.4178, "mean_token_accuracy": 0.8706060647964478, "num_tokens": 885792485.0, "step": 23218 }, { "epoch": 2.9536954585930544, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.729377746582031, "learning_rate": 1e-06, "loss": 0.3924, "mean_token_accuracy": 0.8761656284332275, "num_tokens": 885834809.0, "step": 23219 }, { "epoch": 2.953822668871645, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.67154598236084, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8797023296356201, "num_tokens": 885870555.0, "step": 23220 }, { "epoch": 2.9539498791502354, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.70722484588623, "learning_rate": 1e-06, "loss": 0.3913, "mean_token_accuracy": 0.8769645094871521, "num_tokens": 885911911.0, "step": 23221 }, { "epoch": 2.954077089428826, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.712063789367676, "learning_rate": 1e-06, "loss": 0.3593, "mean_token_accuracy": 0.8863735198974609, "num_tokens": 885941094.0, "step": 23222 }, { "epoch": 2.9542042997074165, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.677064895629883, "learning_rate": 1e-06, "loss": 0.4098, "mean_token_accuracy": 0.8730584979057312, "num_tokens": 885978094.0, "step": 23223 }, { "epoch": 2.954331509986007, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.644482612609863, "learning_rate": 1e-06, "loss": 0.3536, "mean_token_accuracy": 0.8917496204376221, "num_tokens": 886014859.0, "step": 23224 }, { "epoch": 2.954458720264597, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.738285064697266, "learning_rate": 1e-06, "loss": 0.3818, "mean_token_accuracy": 0.8810399770736694, "num_tokens": 886048317.0, "step": 23225 }, { "epoch": 2.954585930543188, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.641470909118652, "learning_rate": 1e-06, "loss": 0.3897, "mean_token_accuracy": 0.877331018447876, "num_tokens": 886081797.0, "step": 23226 }, { "epoch": 2.954713140821778, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 14.718338012695312, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8613513708114624, "num_tokens": 886124111.0, "step": 23227 }, { "epoch": 2.954840351100369, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.653071403503418, "learning_rate": 1e-06, "loss": 0.381, "mean_token_accuracy": 0.8848661780357361, "num_tokens": 886162823.0, "step": 23228 }, { "epoch": 2.954967561378959, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.645692825317383, "learning_rate": 1e-06, "loss": 0.3928, "mean_token_accuracy": 0.8780109882354736, "num_tokens": 886204706.0, "step": 23229 }, { "epoch": 2.95509477165755, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.686180114746094, "learning_rate": 1e-06, "loss": 0.3764, "mean_token_accuracy": 0.8829612731933594, "num_tokens": 886241461.0, "step": 23230 }, { "epoch": 2.9552219819361403, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.667378425598145, "learning_rate": 1e-06, "loss": 0.4118, "mean_token_accuracy": 0.871036171913147, "num_tokens": 886279276.0, "step": 23231 }, { "epoch": 2.955349192214731, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.721759796142578, "learning_rate": 1e-06, "loss": 0.3891, "mean_token_accuracy": 0.8777267932891846, "num_tokens": 886317703.0, "step": 23232 }, { "epoch": 2.9554764024933213, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.639531135559082, "learning_rate": 1e-06, "loss": 0.3802, "mean_token_accuracy": 0.8809922933578491, "num_tokens": 886355235.0, "step": 23233 }, { "epoch": 2.955603612771912, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.71384334564209, "learning_rate": 1e-06, "loss": 0.3257, "mean_token_accuracy": 0.8982068300247192, "num_tokens": 886387354.0, "step": 23234 }, { "epoch": 2.9557308230505024, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.74601936340332, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8721672892570496, "num_tokens": 886419742.0, "step": 23235 }, { "epoch": 2.955858033329093, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.684531211853027, "learning_rate": 1e-06, "loss": 0.3929, "mean_token_accuracy": 0.8748961687088013, "num_tokens": 886455218.0, "step": 23236 }, { "epoch": 2.9559852436076834, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.722294807434082, "learning_rate": 1e-06, "loss": 0.394, "mean_token_accuracy": 0.876865565776825, "num_tokens": 886494454.0, "step": 23237 }, { "epoch": 2.956112453886274, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.661775588989258, "learning_rate": 1e-06, "loss": 0.4461, "mean_token_accuracy": 0.8532278537750244, "num_tokens": 886532854.0, "step": 23238 }, { "epoch": 2.9562396641648645, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.745450973510742, "learning_rate": 1e-06, "loss": 0.3984, "mean_token_accuracy": 0.8732573986053467, "num_tokens": 886564288.0, "step": 23239 }, { "epoch": 2.956366874443455, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.737106323242188, "learning_rate": 1e-06, "loss": 0.3824, "mean_token_accuracy": 0.8775883913040161, "num_tokens": 886597155.0, "step": 23240 }, { "epoch": 2.9564940847220456, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.698559761047363, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.8781981468200684, "num_tokens": 886632015.0, "step": 23241 }, { "epoch": 2.956621295000636, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.710251808166504, "learning_rate": 1e-06, "loss": 0.3407, "mean_token_accuracy": 0.8924211263656616, "num_tokens": 886665736.0, "step": 23242 }, { "epoch": 2.9567485052792266, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.739606857299805, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8742077350616455, "num_tokens": 886702652.0, "step": 23243 }, { "epoch": 2.956875715557817, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.748472213745117, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8729100227355957, "num_tokens": 886743573.0, "step": 23244 }, { "epoch": 2.9570029258364077, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.743805885314941, "learning_rate": 1e-06, "loss": 0.3829, "mean_token_accuracy": 0.8803633451461792, "num_tokens": 886777932.0, "step": 23245 }, { "epoch": 2.957130136114998, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.721503257751465, "learning_rate": 1e-06, "loss": 0.4409, "mean_token_accuracy": 0.8628066778182983, "num_tokens": 886813211.0, "step": 23246 }, { "epoch": 2.9572573463935887, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.708627700805664, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8795508146286011, "num_tokens": 886849416.0, "step": 23247 }, { "epoch": 2.9573845566721793, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.705828666687012, "learning_rate": 1e-06, "loss": 0.4487, "mean_token_accuracy": 0.860569953918457, "num_tokens": 886887983.0, "step": 23248 }, { "epoch": 2.95751176695077, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.720897674560547, "learning_rate": 1e-06, "loss": 0.448, "mean_token_accuracy": 0.8593900799751282, "num_tokens": 886930867.0, "step": 23249 }, { "epoch": 2.95763897722936, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.61390495300293, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8726018071174622, "num_tokens": 886971862.0, "step": 23250 }, { "epoch": 2.957766187507951, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.738234519958496, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8818114399909973, "num_tokens": 887005238.0, "step": 23251 }, { "epoch": 2.957893397786541, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.699167251586914, "learning_rate": 1e-06, "loss": 0.4067, "mean_token_accuracy": 0.8687788844108582, "num_tokens": 887035588.0, "step": 23252 }, { "epoch": 2.958020608065132, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.690059661865234, "learning_rate": 1e-06, "loss": 0.3959, "mean_token_accuracy": 0.8765814304351807, "num_tokens": 887079962.0, "step": 23253 }, { "epoch": 2.958147818343722, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.711812973022461, "learning_rate": 1e-06, "loss": 0.3345, "mean_token_accuracy": 0.8944162726402283, "num_tokens": 887114572.0, "step": 23254 }, { "epoch": 2.958275028622313, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.659019470214844, "learning_rate": 1e-06, "loss": 0.3812, "mean_token_accuracy": 0.8795351386070251, "num_tokens": 887149396.0, "step": 23255 }, { "epoch": 2.958402238900903, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.691823959350586, "learning_rate": 1e-06, "loss": 0.4084, "mean_token_accuracy": 0.8695688247680664, "num_tokens": 887191916.0, "step": 23256 }, { "epoch": 2.9585294491794936, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.667937278747559, "learning_rate": 1e-06, "loss": 0.378, "mean_token_accuracy": 0.884476363658905, "num_tokens": 887228263.0, "step": 23257 }, { "epoch": 2.958656659458084, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.724746704101562, "learning_rate": 1e-06, "loss": 0.3562, "mean_token_accuracy": 0.890143632888794, "num_tokens": 887258688.0, "step": 23258 }, { "epoch": 2.9587838697366746, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.68465518951416, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.8713921904563904, "num_tokens": 887301565.0, "step": 23259 }, { "epoch": 2.958911080015265, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.673283576965332, "learning_rate": 1e-06, "loss": 0.4061, "mean_token_accuracy": 0.8725029230117798, "num_tokens": 887337511.0, "step": 23260 }, { "epoch": 2.9590382902938557, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.692408561706543, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.8800849914550781, "num_tokens": 887370694.0, "step": 23261 }, { "epoch": 2.959165500572446, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.668341636657715, "learning_rate": 1e-06, "loss": 0.4095, "mean_token_accuracy": 0.8738936185836792, "num_tokens": 887412131.0, "step": 23262 }, { "epoch": 2.9592927108510367, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.661796569824219, "learning_rate": 1e-06, "loss": 0.4138, "mean_token_accuracy": 0.8707283735275269, "num_tokens": 887447179.0, "step": 23263 }, { "epoch": 2.9594199211296273, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.699712753295898, "learning_rate": 1e-06, "loss": 0.4144, "mean_token_accuracy": 0.8710527420043945, "num_tokens": 887486079.0, "step": 23264 }, { "epoch": 2.959547131408218, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.631843566894531, "learning_rate": 1e-06, "loss": 0.3759, "mean_token_accuracy": 0.8841805458068848, "num_tokens": 887519092.0, "step": 23265 }, { "epoch": 2.9596743416868083, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.721508026123047, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8784083724021912, "num_tokens": 887552008.0, "step": 23266 }, { "epoch": 2.959801551965399, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.651930809020996, "learning_rate": 1e-06, "loss": 0.4068, "mean_token_accuracy": 0.8709018230438232, "num_tokens": 887585827.0, "step": 23267 }, { "epoch": 2.9599287622439894, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.627972602844238, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.873836100101471, "num_tokens": 887628041.0, "step": 23268 }, { "epoch": 2.96005597252258, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.649420738220215, "learning_rate": 1e-06, "loss": 0.3786, "mean_token_accuracy": 0.8793770670890808, "num_tokens": 887667210.0, "step": 23269 }, { "epoch": 2.9601831828011704, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.649138450622559, "learning_rate": 1e-06, "loss": 0.3712, "mean_token_accuracy": 0.8811037540435791, "num_tokens": 887705429.0, "step": 23270 }, { "epoch": 2.960310393079761, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.64024829864502, "learning_rate": 1e-06, "loss": 0.4627, "mean_token_accuracy": 0.8573279976844788, "num_tokens": 887753188.0, "step": 23271 }, { "epoch": 2.9604376033583515, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.701691627502441, "learning_rate": 1e-06, "loss": 0.441, "mean_token_accuracy": 0.85989910364151, "num_tokens": 887795907.0, "step": 23272 }, { "epoch": 2.960564813636942, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.624198913574219, "learning_rate": 1e-06, "loss": 0.4058, "mean_token_accuracy": 0.8748664259910583, "num_tokens": 887834859.0, "step": 23273 }, { "epoch": 2.9606920239155325, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.661465644836426, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8705506324768066, "num_tokens": 887879041.0, "step": 23274 }, { "epoch": 2.9608192341941226, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.634410858154297, "learning_rate": 1e-06, "loss": 0.4075, "mean_token_accuracy": 0.8718355894088745, "num_tokens": 887913752.0, "step": 23275 }, { "epoch": 2.9609464444727136, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.644524574279785, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8811558485031128, "num_tokens": 887950946.0, "step": 23276 }, { "epoch": 2.9610736547513037, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.645012855529785, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8731251358985901, "num_tokens": 887993155.0, "step": 23277 }, { "epoch": 2.9612008650298947, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.621757507324219, "learning_rate": 1e-06, "loss": 0.4127, "mean_token_accuracy": 0.8739993572235107, "num_tokens": 888027831.0, "step": 23278 }, { "epoch": 2.9613280753084847, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.727041244506836, "learning_rate": 1e-06, "loss": 0.3647, "mean_token_accuracy": 0.8854434490203857, "num_tokens": 888066941.0, "step": 23279 }, { "epoch": 2.9614552855870753, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.639148712158203, "learning_rate": 1e-06, "loss": 0.351, "mean_token_accuracy": 0.8888791799545288, "num_tokens": 888099739.0, "step": 23280 }, { "epoch": 2.961582495865666, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.651649475097656, "learning_rate": 1e-06, "loss": 0.3898, "mean_token_accuracy": 0.8760518431663513, "num_tokens": 888139675.0, "step": 23281 }, { "epoch": 2.9617097061442563, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.752861976623535, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8628530502319336, "num_tokens": 888179374.0, "step": 23282 }, { "epoch": 2.961836916422847, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.658185005187988, "learning_rate": 1e-06, "loss": 0.4301, "mean_token_accuracy": 0.86323082447052, "num_tokens": 888214139.0, "step": 23283 }, { "epoch": 2.9619641267014374, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.67331314086914, "learning_rate": 1e-06, "loss": 0.3981, "mean_token_accuracy": 0.8735618591308594, "num_tokens": 888246365.0, "step": 23284 }, { "epoch": 2.962091336980028, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.691654205322266, "learning_rate": 1e-06, "loss": 0.4206, "mean_token_accuracy": 0.8692277669906616, "num_tokens": 888290903.0, "step": 23285 }, { "epoch": 2.9622185472586184, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.67405891418457, "learning_rate": 1e-06, "loss": 0.4091, "mean_token_accuracy": 0.8724963665008545, "num_tokens": 888331329.0, "step": 23286 }, { "epoch": 2.962345757537209, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.747775077819824, "learning_rate": 1e-06, "loss": 0.4145, "mean_token_accuracy": 0.8682724237442017, "num_tokens": 888372411.0, "step": 23287 }, { "epoch": 2.9624729678157995, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.684215545654297, "learning_rate": 1e-06, "loss": 0.3993, "mean_token_accuracy": 0.8742748498916626, "num_tokens": 888409778.0, "step": 23288 }, { "epoch": 2.96260017809439, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.664998054504395, "learning_rate": 1e-06, "loss": 0.3784, "mean_token_accuracy": 0.8804835081100464, "num_tokens": 888453593.0, "step": 23289 }, { "epoch": 2.9627273883729806, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.651448249816895, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.886768102645874, "num_tokens": 888488245.0, "step": 23290 }, { "epoch": 2.962854598651571, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.626166343688965, "learning_rate": 1e-06, "loss": 0.3936, "mean_token_accuracy": 0.8742125034332275, "num_tokens": 888524306.0, "step": 23291 }, { "epoch": 2.9629818089301616, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.672476768493652, "learning_rate": 1e-06, "loss": 0.3646, "mean_token_accuracy": 0.8876591920852661, "num_tokens": 888563791.0, "step": 23292 }, { "epoch": 2.963109019208752, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.645286560058594, "learning_rate": 1e-06, "loss": 0.3851, "mean_token_accuracy": 0.8790243864059448, "num_tokens": 888605021.0, "step": 23293 }, { "epoch": 2.9632362294873427, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.62948226928711, "learning_rate": 1e-06, "loss": 0.3919, "mean_token_accuracy": 0.875752329826355, "num_tokens": 888647382.0, "step": 23294 }, { "epoch": 2.963363439765933, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.686501502990723, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8747488260269165, "num_tokens": 888682174.0, "step": 23295 }, { "epoch": 2.9634906500445237, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.63300609588623, "learning_rate": 1e-06, "loss": 0.3801, "mean_token_accuracy": 0.8765292167663574, "num_tokens": 888724401.0, "step": 23296 }, { "epoch": 2.9636178603231143, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.728137016296387, "learning_rate": 1e-06, "loss": 0.3799, "mean_token_accuracy": 0.8816410303115845, "num_tokens": 888766230.0, "step": 23297 }, { "epoch": 2.963745070601705, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.651530265808105, "learning_rate": 1e-06, "loss": 0.4604, "mean_token_accuracy": 0.8535385727882385, "num_tokens": 888811865.0, "step": 23298 }, { "epoch": 2.9638722808802953, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.71788215637207, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8753669261932373, "num_tokens": 888849549.0, "step": 23299 }, { "epoch": 2.9639994911588854, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.70777416229248, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8769622445106506, "num_tokens": 888881305.0, "step": 23300 }, { "epoch": 2.9641267014374764, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.671425819396973, "learning_rate": 1e-06, "loss": 0.3875, "mean_token_accuracy": 0.8797386288642883, "num_tokens": 888917862.0, "step": 23301 }, { "epoch": 2.9642539117160664, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.717251777648926, "learning_rate": 1e-06, "loss": 0.3958, "mean_token_accuracy": 0.875456690788269, "num_tokens": 888953883.0, "step": 23302 }, { "epoch": 2.9643811219946574, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.674509048461914, "learning_rate": 1e-06, "loss": 0.3476, "mean_token_accuracy": 0.89238440990448, "num_tokens": 888991052.0, "step": 23303 }, { "epoch": 2.9645083322732475, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.665024757385254, "learning_rate": 1e-06, "loss": 0.3653, "mean_token_accuracy": 0.8878753185272217, "num_tokens": 889022523.0, "step": 23304 }, { "epoch": 2.964635542551838, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.67103099822998, "learning_rate": 1e-06, "loss": 0.42, "mean_token_accuracy": 0.8664062023162842, "num_tokens": 889063470.0, "step": 23305 }, { "epoch": 2.9647627528304286, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.688648223876953, "learning_rate": 1e-06, "loss": 0.3708, "mean_token_accuracy": 0.8834630250930786, "num_tokens": 889101160.0, "step": 23306 }, { "epoch": 2.964889963109019, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.70266342163086, "learning_rate": 1e-06, "loss": 0.4047, "mean_token_accuracy": 0.8718418478965759, "num_tokens": 889139128.0, "step": 23307 }, { "epoch": 2.9650171733876096, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.732144355773926, "learning_rate": 1e-06, "loss": 0.4612, "mean_token_accuracy": 0.8529521822929382, "num_tokens": 889175574.0, "step": 23308 }, { "epoch": 2.9651443836662, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.691851615905762, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8731296062469482, "num_tokens": 889216520.0, "step": 23309 }, { "epoch": 2.9652715939447907, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.65284252166748, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8863608241081238, "num_tokens": 889256746.0, "step": 23310 }, { "epoch": 2.965398804223381, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.720697402954102, "learning_rate": 1e-06, "loss": 0.417, "mean_token_accuracy": 0.8706169128417969, "num_tokens": 889298868.0, "step": 23311 }, { "epoch": 2.9655260145019717, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.726778984069824, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8713334798812866, "num_tokens": 889337758.0, "step": 23312 }, { "epoch": 2.9656532247805623, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.726953506469727, "learning_rate": 1e-06, "loss": 0.389, "mean_token_accuracy": 0.879631757736206, "num_tokens": 889375308.0, "step": 23313 }, { "epoch": 2.965780435059153, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.743678092956543, "learning_rate": 1e-06, "loss": 0.3687, "mean_token_accuracy": 0.8833262920379639, "num_tokens": 889413735.0, "step": 23314 }, { "epoch": 2.9659076453377433, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.658082962036133, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8779306411743164, "num_tokens": 889451585.0, "step": 23315 }, { "epoch": 2.966034855616334, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.779544830322266, "learning_rate": 1e-06, "loss": 0.4518, "mean_token_accuracy": 0.8587232828140259, "num_tokens": 889487405.0, "step": 23316 }, { "epoch": 2.9661620658949244, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.697795867919922, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8637938499450684, "num_tokens": 889523051.0, "step": 23317 }, { "epoch": 2.966289276173515, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.753753662109375, "learning_rate": 1e-06, "loss": 0.4435, "mean_token_accuracy": 0.8572176694869995, "num_tokens": 889559822.0, "step": 23318 }, { "epoch": 2.9664164864521054, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.704662322998047, "learning_rate": 1e-06, "loss": 0.3939, "mean_token_accuracy": 0.8752391338348389, "num_tokens": 889591598.0, "step": 23319 }, { "epoch": 2.966543696730696, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.76733684539795, "learning_rate": 1e-06, "loss": 0.364, "mean_token_accuracy": 0.8849923610687256, "num_tokens": 889629787.0, "step": 23320 }, { "epoch": 2.9666709070092865, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.722549438476562, "learning_rate": 1e-06, "loss": 0.3777, "mean_token_accuracy": 0.8818490505218506, "num_tokens": 889668730.0, "step": 23321 }, { "epoch": 2.966798117287877, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.731588363647461, "learning_rate": 1e-06, "loss": 0.4033, "mean_token_accuracy": 0.8733259439468384, "num_tokens": 889705215.0, "step": 23322 }, { "epoch": 2.966925327566467, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.768013000488281, "learning_rate": 1e-06, "loss": 0.398, "mean_token_accuracy": 0.8760731220245361, "num_tokens": 889738528.0, "step": 23323 }, { "epoch": 2.967052537845058, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.733304977416992, "learning_rate": 1e-06, "loss": 0.3945, "mean_token_accuracy": 0.8800267577171326, "num_tokens": 889771694.0, "step": 23324 }, { "epoch": 2.967179748123648, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.758443832397461, "learning_rate": 1e-06, "loss": 0.3654, "mean_token_accuracy": 0.8881708383560181, "num_tokens": 889806096.0, "step": 23325 }, { "epoch": 2.967306958402239, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.683304786682129, "learning_rate": 1e-06, "loss": 0.4105, "mean_token_accuracy": 0.8739179372787476, "num_tokens": 889848514.0, "step": 23326 }, { "epoch": 2.967434168680829, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.705286979675293, "learning_rate": 1e-06, "loss": 0.3416, "mean_token_accuracy": 0.8951806426048279, "num_tokens": 889889689.0, "step": 23327 }, { "epoch": 2.96756137895942, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.705658912658691, "learning_rate": 1e-06, "loss": 0.3886, "mean_token_accuracy": 0.8782354593276978, "num_tokens": 889925550.0, "step": 23328 }, { "epoch": 2.9676885892380103, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.745783805847168, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8628302216529846, "num_tokens": 889961797.0, "step": 23329 }, { "epoch": 2.967815799516601, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.684809684753418, "learning_rate": 1e-06, "loss": 0.3821, "mean_token_accuracy": 0.8812058568000793, "num_tokens": 889997935.0, "step": 23330 }, { "epoch": 2.9679430097951913, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.724446296691895, "learning_rate": 1e-06, "loss": 0.3589, "mean_token_accuracy": 0.8864650726318359, "num_tokens": 890037900.0, "step": 23331 }, { "epoch": 2.968070220073782, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.755892753601074, "learning_rate": 1e-06, "loss": 0.3798, "mean_token_accuracy": 0.8806830644607544, "num_tokens": 890074814.0, "step": 23332 }, { "epoch": 2.9681974303523724, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.688788414001465, "learning_rate": 1e-06, "loss": 0.3649, "mean_token_accuracy": 0.8879413604736328, "num_tokens": 890107091.0, "step": 23333 }, { "epoch": 2.968324640630963, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.75006103515625, "learning_rate": 1e-06, "loss": 0.3771, "mean_token_accuracy": 0.8839729428291321, "num_tokens": 890140558.0, "step": 23334 }, { "epoch": 2.9684518509095534, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.79023551940918, "learning_rate": 1e-06, "loss": 0.3935, "mean_token_accuracy": 0.87631756067276, "num_tokens": 890174563.0, "step": 23335 }, { "epoch": 2.968579061188144, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.73473834991455, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.873511552810669, "num_tokens": 890210584.0, "step": 23336 }, { "epoch": 2.9687062714667345, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.719574928283691, "learning_rate": 1e-06, "loss": 0.3598, "mean_token_accuracy": 0.887760579586029, "num_tokens": 890251955.0, "step": 23337 }, { "epoch": 2.968833481745325, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.724563598632812, "learning_rate": 1e-06, "loss": 0.4024, "mean_token_accuracy": 0.8705360293388367, "num_tokens": 890289095.0, "step": 23338 }, { "epoch": 2.9689606920239155, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.695732116699219, "learning_rate": 1e-06, "loss": 0.3971, "mean_token_accuracy": 0.8775554895401001, "num_tokens": 890327864.0, "step": 23339 }, { "epoch": 2.969087902302506, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.774213790893555, "learning_rate": 1e-06, "loss": 0.3938, "mean_token_accuracy": 0.8752709031105042, "num_tokens": 890364244.0, "step": 23340 }, { "epoch": 2.9692151125810966, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.750128746032715, "learning_rate": 1e-06, "loss": 0.4479, "mean_token_accuracy": 0.8611751198768616, "num_tokens": 890397652.0, "step": 23341 }, { "epoch": 2.969342322859687, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.727198600769043, "learning_rate": 1e-06, "loss": 0.3817, "mean_token_accuracy": 0.8801059722900391, "num_tokens": 890435598.0, "step": 23342 }, { "epoch": 2.9694695331382777, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.68867301940918, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8790344595909119, "num_tokens": 890473445.0, "step": 23343 }, { "epoch": 2.969596743416868, "ewc_loss": 0.0400390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.7293701171875, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8797703981399536, "num_tokens": 890514263.0, "step": 23344 }, { "epoch": 2.9697239536954587, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.750131607055664, "learning_rate": 1e-06, "loss": 0.3719, "mean_token_accuracy": 0.8807166814804077, "num_tokens": 890554123.0, "step": 23345 }, { "epoch": 2.9698511639740492, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.68056869506836, "learning_rate": 1e-06, "loss": 0.4071, "mean_token_accuracy": 0.8753229975700378, "num_tokens": 890593938.0, "step": 23346 }, { "epoch": 2.9699783742526398, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.71651840209961, "learning_rate": 1e-06, "loss": 0.4274, "mean_token_accuracy": 0.865643322467804, "num_tokens": 890632260.0, "step": 23347 }, { "epoch": 2.97010558453123, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.68443489074707, "learning_rate": 1e-06, "loss": 0.399, "mean_token_accuracy": 0.872870922088623, "num_tokens": 890666852.0, "step": 23348 }, { "epoch": 2.970232794809821, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.729140281677246, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8718333840370178, "num_tokens": 890707959.0, "step": 23349 }, { "epoch": 2.970360005088411, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.66480827331543, "learning_rate": 1e-06, "loss": 0.4745, "mean_token_accuracy": 0.8545941114425659, "num_tokens": 890748729.0, "step": 23350 }, { "epoch": 2.970487215367002, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.72320556640625, "learning_rate": 1e-06, "loss": 0.3569, "mean_token_accuracy": 0.8873570561408997, "num_tokens": 890782116.0, "step": 23351 }, { "epoch": 2.970614425645592, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.688803672790527, "learning_rate": 1e-06, "loss": 0.4341, "mean_token_accuracy": 0.8625951409339905, "num_tokens": 890827309.0, "step": 23352 }, { "epoch": 2.970741635924183, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.692269325256348, "learning_rate": 1e-06, "loss": 0.44, "mean_token_accuracy": 0.865470290184021, "num_tokens": 890863879.0, "step": 23353 }, { "epoch": 2.970868846202773, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.706338882446289, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.878150224685669, "num_tokens": 890901184.0, "step": 23354 }, { "epoch": 2.9709960564813636, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.716114044189453, "learning_rate": 1e-06, "loss": 0.3643, "mean_token_accuracy": 0.8853800296783447, "num_tokens": 890937779.0, "step": 23355 }, { "epoch": 2.971123266759954, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.693984031677246, "learning_rate": 1e-06, "loss": 0.3535, "mean_token_accuracy": 0.8881466388702393, "num_tokens": 890978055.0, "step": 23356 }, { "epoch": 2.9712504770385446, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.783564567565918, "learning_rate": 1e-06, "loss": 0.5208, "mean_token_accuracy": 0.8457044363021851, "num_tokens": 891010985.0, "step": 23357 }, { "epoch": 2.971377687317135, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.748616218566895, "learning_rate": 1e-06, "loss": 0.3838, "mean_token_accuracy": 0.880448043346405, "num_tokens": 891045904.0, "step": 23358 }, { "epoch": 2.9715048975957257, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.703661918640137, "learning_rate": 1e-06, "loss": 0.3705, "mean_token_accuracy": 0.8790302276611328, "num_tokens": 891086240.0, "step": 23359 }, { "epoch": 2.971632107874316, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.653932571411133, "learning_rate": 1e-06, "loss": 0.4044, "mean_token_accuracy": 0.8731173276901245, "num_tokens": 891126482.0, "step": 23360 }, { "epoch": 2.9717593181529067, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.678206443786621, "learning_rate": 1e-06, "loss": 0.4265, "mean_token_accuracy": 0.8651093244552612, "num_tokens": 891165982.0, "step": 23361 }, { "epoch": 2.9718865284314973, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.668367385864258, "learning_rate": 1e-06, "loss": 0.4338, "mean_token_accuracy": 0.8615936040878296, "num_tokens": 891199101.0, "step": 23362 }, { "epoch": 2.972013738710088, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.774410247802734, "learning_rate": 1e-06, "loss": 0.408, "mean_token_accuracy": 0.8745230436325073, "num_tokens": 891231511.0, "step": 23363 }, { "epoch": 2.9721409489886783, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.736825942993164, "learning_rate": 1e-06, "loss": 0.4385, "mean_token_accuracy": 0.8639183044433594, "num_tokens": 891262369.0, "step": 23364 }, { "epoch": 2.972268159267269, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.716778755187988, "learning_rate": 1e-06, "loss": 0.4366, "mean_token_accuracy": 0.8597174286842346, "num_tokens": 891304206.0, "step": 23365 }, { "epoch": 2.9723953695458594, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.679341316223145, "learning_rate": 1e-06, "loss": 0.4346, "mean_token_accuracy": 0.8629420399665833, "num_tokens": 891348006.0, "step": 23366 }, { "epoch": 2.97252257982445, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.654000282287598, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.8822858333587646, "num_tokens": 891386851.0, "step": 23367 }, { "epoch": 2.9726497901030404, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.648653030395508, "learning_rate": 1e-06, "loss": 0.3714, "mean_token_accuracy": 0.884617805480957, "num_tokens": 891423582.0, "step": 23368 }, { "epoch": 2.972777000381631, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.64244270324707, "learning_rate": 1e-06, "loss": 0.3437, "mean_token_accuracy": 0.8896596431732178, "num_tokens": 891463773.0, "step": 23369 }, { "epoch": 2.9729042106602215, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.66973876953125, "learning_rate": 1e-06, "loss": 0.3623, "mean_token_accuracy": 0.8860530257225037, "num_tokens": 891498117.0, "step": 23370 }, { "epoch": 2.973031420938812, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.764159202575684, "learning_rate": 1e-06, "loss": 0.3703, "mean_token_accuracy": 0.883026123046875, "num_tokens": 891529055.0, "step": 23371 }, { "epoch": 2.9731586312174025, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.659236907958984, "learning_rate": 1e-06, "loss": 0.3674, "mean_token_accuracy": 0.8837292194366455, "num_tokens": 891567370.0, "step": 23372 }, { "epoch": 2.9732858414959926, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.701740264892578, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8780086040496826, "num_tokens": 891602737.0, "step": 23373 }, { "epoch": 2.9734130517745836, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.711681365966797, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8680348992347717, "num_tokens": 891639713.0, "step": 23374 }, { "epoch": 2.9735402620531737, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.714415550231934, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.887927770614624, "num_tokens": 891673931.0, "step": 23375 }, { "epoch": 2.9736674723317646, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.718311309814453, "learning_rate": 1e-06, "loss": 0.4062, "mean_token_accuracy": 0.8749321103096008, "num_tokens": 891708015.0, "step": 23376 }, { "epoch": 2.9737946826103547, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.691311836242676, "learning_rate": 1e-06, "loss": 0.4099, "mean_token_accuracy": 0.8718978762626648, "num_tokens": 891746023.0, "step": 23377 }, { "epoch": 2.9739218928889453, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.741114616394043, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8730987310409546, "num_tokens": 891777767.0, "step": 23378 }, { "epoch": 2.974049103167536, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.725046157836914, "learning_rate": 1e-06, "loss": 0.3827, "mean_token_accuracy": 0.8779782056808472, "num_tokens": 891810816.0, "step": 23379 }, { "epoch": 2.9741763134461263, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.719416618347168, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8687206506729126, "num_tokens": 891848810.0, "step": 23380 }, { "epoch": 2.974303523724717, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.707889556884766, "learning_rate": 1e-06, "loss": 0.3932, "mean_token_accuracy": 0.8789858818054199, "num_tokens": 891892553.0, "step": 23381 }, { "epoch": 2.9744307340033074, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.71343994140625, "learning_rate": 1e-06, "loss": 0.3606, "mean_token_accuracy": 0.8856251239776611, "num_tokens": 891927539.0, "step": 23382 }, { "epoch": 2.974557944281898, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.7727632522583, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8797595500946045, "num_tokens": 891959577.0, "step": 23383 }, { "epoch": 2.9746851545604884, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.66974925994873, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8491575717926025, "num_tokens": 892000799.0, "step": 23384 }, { "epoch": 2.974812364839079, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.74315357208252, "learning_rate": 1e-06, "loss": 0.4173, "mean_token_accuracy": 0.8689943552017212, "num_tokens": 892042200.0, "step": 23385 }, { "epoch": 2.9749395751176695, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.679932594299316, "learning_rate": 1e-06, "loss": 0.4279, "mean_token_accuracy": 0.8660575151443481, "num_tokens": 892084838.0, "step": 23386 }, { "epoch": 2.97506678539626, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.702618598937988, "learning_rate": 1e-06, "loss": 0.4444, "mean_token_accuracy": 0.8594812154769897, "num_tokens": 892126442.0, "step": 23387 }, { "epoch": 2.9751939956748505, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.691856384277344, "learning_rate": 1e-06, "loss": 0.415, "mean_token_accuracy": 0.8656477332115173, "num_tokens": 892165019.0, "step": 23388 }, { "epoch": 2.975321205953441, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.735515594482422, "learning_rate": 1e-06, "loss": 0.3946, "mean_token_accuracy": 0.8768952488899231, "num_tokens": 892202291.0, "step": 23389 }, { "epoch": 2.9754484162320316, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.683621406555176, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8667969703674316, "num_tokens": 892235159.0, "step": 23390 }, { "epoch": 2.975575626510622, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.756281852722168, "learning_rate": 1e-06, "loss": 0.4393, "mean_token_accuracy": 0.8628882169723511, "num_tokens": 892271967.0, "step": 23391 }, { "epoch": 2.9757028367892127, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.681818962097168, "learning_rate": 1e-06, "loss": 0.4472, "mean_token_accuracy": 0.8613916635513306, "num_tokens": 892306230.0, "step": 23392 }, { "epoch": 2.975830047067803, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.77896499633789, "learning_rate": 1e-06, "loss": 0.3833, "mean_token_accuracy": 0.8769454956054688, "num_tokens": 892342118.0, "step": 23393 }, { "epoch": 2.9759572573463937, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.6476411819458, "learning_rate": 1e-06, "loss": 0.3845, "mean_token_accuracy": 0.8800469040870667, "num_tokens": 892380764.0, "step": 23394 }, { "epoch": 2.9760844676249842, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.740934371948242, "learning_rate": 1e-06, "loss": 0.41, "mean_token_accuracy": 0.8721524477005005, "num_tokens": 892421113.0, "step": 23395 }, { "epoch": 2.9762116779035748, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.738262176513672, "learning_rate": 1e-06, "loss": 0.38, "mean_token_accuracy": 0.8815982341766357, "num_tokens": 892457512.0, "step": 23396 }, { "epoch": 2.9763388881821653, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.663086891174316, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8804206252098083, "num_tokens": 892493599.0, "step": 23397 }, { "epoch": 2.9764660984607554, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.844099044799805, "learning_rate": 1e-06, "loss": 0.4168, "mean_token_accuracy": 0.866529107093811, "num_tokens": 892528474.0, "step": 23398 }, { "epoch": 2.9765933087393464, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.73513412475586, "learning_rate": 1e-06, "loss": 0.3731, "mean_token_accuracy": 0.8807661533355713, "num_tokens": 892569591.0, "step": 23399 }, { "epoch": 2.9767205190179364, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.702961921691895, "learning_rate": 1e-06, "loss": 0.3782, "mean_token_accuracy": 0.8797147274017334, "num_tokens": 892607185.0, "step": 23400 }, { "epoch": 2.9768477292965274, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.696494102478027, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.8761935830116272, "num_tokens": 892641453.0, "step": 23401 }, { "epoch": 2.9769749395751175, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.73278522491455, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8693140745162964, "num_tokens": 892676802.0, "step": 23402 }, { "epoch": 2.977102149853708, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.671150207519531, "learning_rate": 1e-06, "loss": 0.3735, "mean_token_accuracy": 0.8864781260490417, "num_tokens": 892711694.0, "step": 23403 }, { "epoch": 2.9772293601322986, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.728653907775879, "learning_rate": 1e-06, "loss": 0.4196, "mean_token_accuracy": 0.8660331964492798, "num_tokens": 892752662.0, "step": 23404 }, { "epoch": 2.977356570410889, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.682997703552246, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.8767709732055664, "num_tokens": 892790170.0, "step": 23405 }, { "epoch": 2.9774837806894796, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.710064888000488, "learning_rate": 1e-06, "loss": 0.4622, "mean_token_accuracy": 0.8574497103691101, "num_tokens": 892831437.0, "step": 23406 }, { "epoch": 2.97761099096807, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.643019676208496, "learning_rate": 1e-06, "loss": 0.3337, "mean_token_accuracy": 0.8970187902450562, "num_tokens": 892869022.0, "step": 23407 }, { "epoch": 2.9777382012466607, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.764108657836914, "learning_rate": 1e-06, "loss": 0.3974, "mean_token_accuracy": 0.8764952421188354, "num_tokens": 892910738.0, "step": 23408 }, { "epoch": 2.977865411525251, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.673614501953125, "learning_rate": 1e-06, "loss": 0.3809, "mean_token_accuracy": 0.8827692270278931, "num_tokens": 892952055.0, "step": 23409 }, { "epoch": 2.9779926218038417, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.664546966552734, "learning_rate": 1e-06, "loss": 0.4263, "mean_token_accuracy": 0.8676775693893433, "num_tokens": 892985843.0, "step": 23410 }, { "epoch": 2.9781198320824323, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.723137855529785, "learning_rate": 1e-06, "loss": 0.4174, "mean_token_accuracy": 0.8650214672088623, "num_tokens": 893031197.0, "step": 23411 }, { "epoch": 2.978247042361023, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.708762168884277, "learning_rate": 1e-06, "loss": 0.3492, "mean_token_accuracy": 0.889147162437439, "num_tokens": 893068113.0, "step": 23412 }, { "epoch": 2.9783742526396133, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.68873119354248, "learning_rate": 1e-06, "loss": 0.3658, "mean_token_accuracy": 0.8855265974998474, "num_tokens": 893109148.0, "step": 23413 }, { "epoch": 2.978501462918204, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.686623573303223, "learning_rate": 1e-06, "loss": 0.3816, "mean_token_accuracy": 0.8801467418670654, "num_tokens": 893151406.0, "step": 23414 }, { "epoch": 2.9786286731967944, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.720791816711426, "learning_rate": 1e-06, "loss": 0.3826, "mean_token_accuracy": 0.8789361715316772, "num_tokens": 893190209.0, "step": 23415 }, { "epoch": 2.978755883475385, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.718881607055664, "learning_rate": 1e-06, "loss": 0.4109, "mean_token_accuracy": 0.8744546175003052, "num_tokens": 893232949.0, "step": 23416 }, { "epoch": 2.9788830937539754, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.67242431640625, "learning_rate": 1e-06, "loss": 0.3966, "mean_token_accuracy": 0.8752507567405701, "num_tokens": 893273931.0, "step": 23417 }, { "epoch": 2.979010304032566, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.700722694396973, "learning_rate": 1e-06, "loss": 0.4559, "mean_token_accuracy": 0.8557785749435425, "num_tokens": 893317465.0, "step": 23418 }, { "epoch": 2.9791375143111565, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.779403686523438, "learning_rate": 1e-06, "loss": 0.4493, "mean_token_accuracy": 0.8596659898757935, "num_tokens": 893354089.0, "step": 23419 }, { "epoch": 2.979264724589747, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.743484497070312, "learning_rate": 1e-06, "loss": 0.4008, "mean_token_accuracy": 0.8729947805404663, "num_tokens": 893388051.0, "step": 23420 }, { "epoch": 2.979391934868337, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.700222969055176, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8698703050613403, "num_tokens": 893426191.0, "step": 23421 }, { "epoch": 2.979519145146928, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.764923095703125, "learning_rate": 1e-06, "loss": 0.4195, "mean_token_accuracy": 0.8675632476806641, "num_tokens": 893462134.0, "step": 23422 }, { "epoch": 2.979646355425518, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.743331909179688, "learning_rate": 1e-06, "loss": 0.4285, "mean_token_accuracy": 0.8698688745498657, "num_tokens": 893503159.0, "step": 23423 }, { "epoch": 2.979773565704109, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.801709175109863, "learning_rate": 1e-06, "loss": 0.4654, "mean_token_accuracy": 0.8573201894760132, "num_tokens": 893538379.0, "step": 23424 }, { "epoch": 2.979900775982699, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.695948600769043, "learning_rate": 1e-06, "loss": 0.4302, "mean_token_accuracy": 0.8624958992004395, "num_tokens": 893580431.0, "step": 23425 }, { "epoch": 2.98002798626129, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.717970848083496, "learning_rate": 1e-06, "loss": 0.4309, "mean_token_accuracy": 0.868574321269989, "num_tokens": 893622412.0, "step": 23426 }, { "epoch": 2.9801551965398803, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.733075141906738, "learning_rate": 1e-06, "loss": 0.4003, "mean_token_accuracy": 0.8780856132507324, "num_tokens": 893658575.0, "step": 23427 }, { "epoch": 2.980282406818471, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.661846160888672, "learning_rate": 1e-06, "loss": 0.374, "mean_token_accuracy": 0.881401777267456, "num_tokens": 893702812.0, "step": 23428 }, { "epoch": 2.9804096170970613, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.711822509765625, "learning_rate": 1e-06, "loss": 0.4199, "mean_token_accuracy": 0.8678001761436462, "num_tokens": 893742785.0, "step": 23429 }, { "epoch": 2.980536827375652, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.697432518005371, "learning_rate": 1e-06, "loss": 0.3847, "mean_token_accuracy": 0.8803731203079224, "num_tokens": 893782310.0, "step": 23430 }, { "epoch": 2.9806640376542424, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.694116592407227, "learning_rate": 1e-06, "loss": 0.3404, "mean_token_accuracy": 0.8912545442581177, "num_tokens": 893822794.0, "step": 23431 }, { "epoch": 2.980791247932833, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.657071113586426, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8799248933792114, "num_tokens": 893870462.0, "step": 23432 }, { "epoch": 2.9809184582114234, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.702120780944824, "learning_rate": 1e-06, "loss": 0.4073, "mean_token_accuracy": 0.8748809099197388, "num_tokens": 893910654.0, "step": 23433 }, { "epoch": 2.981045668490014, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.682804107666016, "learning_rate": 1e-06, "loss": 0.4006, "mean_token_accuracy": 0.8713291883468628, "num_tokens": 893948819.0, "step": 23434 }, { "epoch": 2.9811728787686045, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.703629493713379, "learning_rate": 1e-06, "loss": 0.3755, "mean_token_accuracy": 0.8799948692321777, "num_tokens": 893983325.0, "step": 23435 }, { "epoch": 2.981300089047195, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.709938049316406, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.8727859258651733, "num_tokens": 894023712.0, "step": 23436 }, { "epoch": 2.9814272993257855, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.668095588684082, "learning_rate": 1e-06, "loss": 0.3681, "mean_token_accuracy": 0.8833584785461426, "num_tokens": 894059719.0, "step": 23437 }, { "epoch": 2.981554509604376, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.711483001708984, "learning_rate": 1e-06, "loss": 0.3949, "mean_token_accuracy": 0.8758466839790344, "num_tokens": 894100150.0, "step": 23438 }, { "epoch": 2.9816817198829666, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.700355529785156, "learning_rate": 1e-06, "loss": 0.4802, "mean_token_accuracy": 0.8544646501541138, "num_tokens": 894134336.0, "step": 23439 }, { "epoch": 2.981808930161557, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.731685638427734, "learning_rate": 1e-06, "loss": 0.4411, "mean_token_accuracy": 0.8613543510437012, "num_tokens": 894168972.0, "step": 23440 }, { "epoch": 2.9819361404401477, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.660261154174805, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8716872334480286, "num_tokens": 894204641.0, "step": 23441 }, { "epoch": 2.982063350718738, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.713183403015137, "learning_rate": 1e-06, "loss": 0.4025, "mean_token_accuracy": 0.8770996928215027, "num_tokens": 894248403.0, "step": 23442 }, { "epoch": 2.9821905609973287, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.680402755737305, "learning_rate": 1e-06, "loss": 0.3397, "mean_token_accuracy": 0.8920802474021912, "num_tokens": 894284035.0, "step": 23443 }, { "epoch": 2.9823177712759192, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.763134956359863, "learning_rate": 1e-06, "loss": 0.3795, "mean_token_accuracy": 0.8784341812133789, "num_tokens": 894321804.0, "step": 23444 }, { "epoch": 2.9824449815545098, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.717208862304688, "learning_rate": 1e-06, "loss": 0.3706, "mean_token_accuracy": 0.8836714029312134, "num_tokens": 894355648.0, "step": 23445 }, { "epoch": 2.9825721918331, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.719696044921875, "learning_rate": 1e-06, "loss": 0.3912, "mean_token_accuracy": 0.875557541847229, "num_tokens": 894392947.0, "step": 23446 }, { "epoch": 2.982699402111691, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.70932674407959, "learning_rate": 1e-06, "loss": 0.4304, "mean_token_accuracy": 0.8659467697143555, "num_tokens": 894433437.0, "step": 23447 }, { "epoch": 2.982826612390281, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.751787185668945, "learning_rate": 1e-06, "loss": 0.3871, "mean_token_accuracy": 0.8797794580459595, "num_tokens": 894481428.0, "step": 23448 }, { "epoch": 2.982953822668872, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.718937873840332, "learning_rate": 1e-06, "loss": 0.3996, "mean_token_accuracy": 0.8742728233337402, "num_tokens": 894518055.0, "step": 23449 }, { "epoch": 2.983081032947462, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.734882354736328, "learning_rate": 1e-06, "loss": 0.4236, "mean_token_accuracy": 0.8701398968696594, "num_tokens": 894562693.0, "step": 23450 }, { "epoch": 2.983208243226053, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.665529251098633, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8749640583992004, "num_tokens": 894604973.0, "step": 23451 }, { "epoch": 2.983335453504643, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.752867698669434, "learning_rate": 1e-06, "loss": 0.376, "mean_token_accuracy": 0.8807352781295776, "num_tokens": 894645672.0, "step": 23452 }, { "epoch": 2.9834626637832335, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.665132522583008, "learning_rate": 1e-06, "loss": 0.4065, "mean_token_accuracy": 0.8734030723571777, "num_tokens": 894686702.0, "step": 23453 }, { "epoch": 2.983589874061824, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.737397193908691, "learning_rate": 1e-06, "loss": 0.3751, "mean_token_accuracy": 0.8831160068511963, "num_tokens": 894722768.0, "step": 23454 }, { "epoch": 2.9837170843404146, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.658077239990234, "learning_rate": 1e-06, "loss": 0.3685, "mean_token_accuracy": 0.8861813545227051, "num_tokens": 894759216.0, "step": 23455 }, { "epoch": 2.983844294619005, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.676304817199707, "learning_rate": 1e-06, "loss": 0.3796, "mean_token_accuracy": 0.8789239525794983, "num_tokens": 894799691.0, "step": 23456 }, { "epoch": 2.9839715048975957, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.685223579406738, "learning_rate": 1e-06, "loss": 0.4162, "mean_token_accuracy": 0.8707746267318726, "num_tokens": 894835869.0, "step": 23457 }, { "epoch": 2.984098715176186, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.766494750976562, "learning_rate": 1e-06, "loss": 0.4573, "mean_token_accuracy": 0.8536026477813721, "num_tokens": 894875706.0, "step": 23458 }, { "epoch": 2.9842259254547767, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.6653413772583, "learning_rate": 1e-06, "loss": 0.3884, "mean_token_accuracy": 0.8802846074104309, "num_tokens": 894920135.0, "step": 23459 }, { "epoch": 2.9843531357333672, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.676800727844238, "learning_rate": 1e-06, "loss": 0.3899, "mean_token_accuracy": 0.8759555816650391, "num_tokens": 894960825.0, "step": 23460 }, { "epoch": 2.9844803460119578, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.755692481994629, "learning_rate": 1e-06, "loss": 0.3972, "mean_token_accuracy": 0.8768994808197021, "num_tokens": 895001935.0, "step": 23461 }, { "epoch": 2.9846075562905483, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.654369354248047, "learning_rate": 1e-06, "loss": 0.3624, "mean_token_accuracy": 0.883769154548645, "num_tokens": 895033559.0, "step": 23462 }, { "epoch": 2.984734766569139, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.694457054138184, "learning_rate": 1e-06, "loss": 0.4029, "mean_token_accuracy": 0.8739520907402039, "num_tokens": 895070958.0, "step": 23463 }, { "epoch": 2.9848619768477294, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.722379684448242, "learning_rate": 1e-06, "loss": 0.4036, "mean_token_accuracy": 0.8728987574577332, "num_tokens": 895111533.0, "step": 23464 }, { "epoch": 2.98498918712632, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.74150276184082, "learning_rate": 1e-06, "loss": 0.4078, "mean_token_accuracy": 0.8709428906440735, "num_tokens": 895149861.0, "step": 23465 }, { "epoch": 2.9851163974049104, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.615521430969238, "learning_rate": 1e-06, "loss": 0.3781, "mean_token_accuracy": 0.8817135095596313, "num_tokens": 895192958.0, "step": 23466 }, { "epoch": 2.985243607683501, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.69330883026123, "learning_rate": 1e-06, "loss": 0.3716, "mean_token_accuracy": 0.8847494125366211, "num_tokens": 895226710.0, "step": 23467 }, { "epoch": 2.9853708179620915, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.727341651916504, "learning_rate": 1e-06, "loss": 0.3872, "mean_token_accuracy": 0.879379153251648, "num_tokens": 895267695.0, "step": 23468 }, { "epoch": 2.985498028240682, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.65562629699707, "learning_rate": 1e-06, "loss": 0.3555, "mean_token_accuracy": 0.891011118888855, "num_tokens": 895303101.0, "step": 23469 }, { "epoch": 2.9856252385192725, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.686339378356934, "learning_rate": 1e-06, "loss": 0.3767, "mean_token_accuracy": 0.8808807134628296, "num_tokens": 895340437.0, "step": 23470 }, { "epoch": 2.9857524487978626, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.65894603729248, "learning_rate": 1e-06, "loss": 0.3587, "mean_token_accuracy": 0.8895134925842285, "num_tokens": 895377695.0, "step": 23471 }, { "epoch": 2.9858796590764536, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.68660831451416, "learning_rate": 1e-06, "loss": 0.4259, "mean_token_accuracy": 0.8663055300712585, "num_tokens": 895418599.0, "step": 23472 }, { "epoch": 2.9860068693550437, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.723416328430176, "learning_rate": 1e-06, "loss": 0.3973, "mean_token_accuracy": 0.8740646839141846, "num_tokens": 895454832.0, "step": 23473 }, { "epoch": 2.9861340796336346, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.73390007019043, "learning_rate": 1e-06, "loss": 0.4585, "mean_token_accuracy": 0.8583160042762756, "num_tokens": 895490607.0, "step": 23474 }, { "epoch": 2.9862612899122247, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.705252647399902, "learning_rate": 1e-06, "loss": 0.4129, "mean_token_accuracy": 0.8686237335205078, "num_tokens": 895531340.0, "step": 23475 }, { "epoch": 2.9863885001908153, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.684396743774414, "learning_rate": 1e-06, "loss": 0.4004, "mean_token_accuracy": 0.8756731748580933, "num_tokens": 895568688.0, "step": 23476 }, { "epoch": 2.986515710469406, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.696465492248535, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8655834794044495, "num_tokens": 895607316.0, "step": 23477 }, { "epoch": 2.9866429207479963, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.75527572631836, "learning_rate": 1e-06, "loss": 0.3459, "mean_token_accuracy": 0.8916996121406555, "num_tokens": 895648694.0, "step": 23478 }, { "epoch": 2.986770131026587, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.695795059204102, "learning_rate": 1e-06, "loss": 0.4141, "mean_token_accuracy": 0.8706032633781433, "num_tokens": 895687996.0, "step": 23479 }, { "epoch": 2.9868973413051774, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.669130325317383, "learning_rate": 1e-06, "loss": 0.3839, "mean_token_accuracy": 0.8795289993286133, "num_tokens": 895726727.0, "step": 23480 }, { "epoch": 2.987024551583768, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.714592933654785, "learning_rate": 1e-06, "loss": 0.3748, "mean_token_accuracy": 0.8796868324279785, "num_tokens": 895760688.0, "step": 23481 }, { "epoch": 2.9871517618623584, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.71201229095459, "learning_rate": 1e-06, "loss": 0.4547, "mean_token_accuracy": 0.8542145490646362, "num_tokens": 895797303.0, "step": 23482 }, { "epoch": 2.987278972140949, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.677645683288574, "learning_rate": 1e-06, "loss": 0.3794, "mean_token_accuracy": 0.8810791969299316, "num_tokens": 895834078.0, "step": 23483 }, { "epoch": 2.9874061824195395, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.745501518249512, "learning_rate": 1e-06, "loss": 0.4339, "mean_token_accuracy": 0.8650112152099609, "num_tokens": 895872640.0, "step": 23484 }, { "epoch": 2.98753339269813, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.73445987701416, "learning_rate": 1e-06, "loss": 0.3622, "mean_token_accuracy": 0.888556957244873, "num_tokens": 895907646.0, "step": 23485 }, { "epoch": 2.9876606029767205, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.689258575439453, "learning_rate": 1e-06, "loss": 0.3723, "mean_token_accuracy": 0.8844928741455078, "num_tokens": 895944183.0, "step": 23486 }, { "epoch": 2.987787813255311, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.775668144226074, "learning_rate": 1e-06, "loss": 0.3848, "mean_token_accuracy": 0.879206657409668, "num_tokens": 895983987.0, "step": 23487 }, { "epoch": 2.9879150235339016, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.7881498336792, "learning_rate": 1e-06, "loss": 0.4282, "mean_token_accuracy": 0.8637651205062866, "num_tokens": 896020928.0, "step": 23488 }, { "epoch": 2.988042233812492, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.686799049377441, "learning_rate": 1e-06, "loss": 0.369, "mean_token_accuracy": 0.8820800185203552, "num_tokens": 896057459.0, "step": 23489 }, { "epoch": 2.9881694440910826, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.738784790039062, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8712775111198425, "num_tokens": 896095049.0, "step": 23490 }, { "epoch": 2.988296654369673, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.725820541381836, "learning_rate": 1e-06, "loss": 0.3662, "mean_token_accuracy": 0.8864495754241943, "num_tokens": 896127433.0, "step": 23491 }, { "epoch": 2.9884238646482637, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.737811088562012, "learning_rate": 1e-06, "loss": 0.383, "mean_token_accuracy": 0.8811132907867432, "num_tokens": 896165662.0, "step": 23492 }, { "epoch": 2.9885510749268542, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.749776840209961, "learning_rate": 1e-06, "loss": 0.4031, "mean_token_accuracy": 0.8738464713096619, "num_tokens": 896207753.0, "step": 23493 }, { "epoch": 2.9886782852054448, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.712472915649414, "learning_rate": 1e-06, "loss": 0.3985, "mean_token_accuracy": 0.8746700286865234, "num_tokens": 896249105.0, "step": 23494 }, { "epoch": 2.9888054954840353, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.767587661743164, "learning_rate": 1e-06, "loss": 0.3933, "mean_token_accuracy": 0.8753902316093445, "num_tokens": 896291728.0, "step": 23495 }, { "epoch": 2.9889327057626254, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.644865989685059, "learning_rate": 1e-06, "loss": 0.4272, "mean_token_accuracy": 0.8656151294708252, "num_tokens": 896331729.0, "step": 23496 }, { "epoch": 2.9890599160412163, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.744331359863281, "learning_rate": 1e-06, "loss": 0.4267, "mean_token_accuracy": 0.8643985390663147, "num_tokens": 896371804.0, "step": 23497 }, { "epoch": 2.9891871263198064, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.776634216308594, "learning_rate": 1e-06, "loss": 0.3965, "mean_token_accuracy": 0.873626172542572, "num_tokens": 896408398.0, "step": 23498 }, { "epoch": 2.9893143365983974, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.682355880737305, "learning_rate": 1e-06, "loss": 0.3636, "mean_token_accuracy": 0.8903191089630127, "num_tokens": 896440496.0, "step": 23499 }, { "epoch": 2.9894415468769875, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.768815994262695, "learning_rate": 1e-06, "loss": 0.3987, "mean_token_accuracy": 0.877084493637085, "num_tokens": 896475212.0, "step": 23500 }, { "epoch": 2.989568757155578, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.724370002746582, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8722161054611206, "num_tokens": 896517122.0, "step": 23501 }, { "epoch": 2.9896959674341685, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.716827392578125, "learning_rate": 1e-06, "loss": 0.3757, "mean_token_accuracy": 0.882206380367279, "num_tokens": 896557259.0, "step": 23502 }, { "epoch": 2.989823177712759, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.695300102233887, "learning_rate": 1e-06, "loss": 0.4244, "mean_token_accuracy": 0.8661074638366699, "num_tokens": 896600237.0, "step": 23503 }, { "epoch": 2.9899503879913496, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.7376070022583, "learning_rate": 1e-06, "loss": 0.3627, "mean_token_accuracy": 0.8864942193031311, "num_tokens": 896640564.0, "step": 23504 }, { "epoch": 2.99007759826994, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.749934196472168, "learning_rate": 1e-06, "loss": 0.3565, "mean_token_accuracy": 0.8850678205490112, "num_tokens": 896674164.0, "step": 23505 }, { "epoch": 2.9902048085485307, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.74758529663086, "learning_rate": 1e-06, "loss": 0.4233, "mean_token_accuracy": 0.8700360655784607, "num_tokens": 896716613.0, "step": 23506 }, { "epoch": 2.990332018827121, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.703758239746094, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8647571206092834, "num_tokens": 896755586.0, "step": 23507 }, { "epoch": 2.9904592291057117, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.72995662689209, "learning_rate": 1e-06, "loss": 0.3982, "mean_token_accuracy": 0.873753011226654, "num_tokens": 896797641.0, "step": 23508 }, { "epoch": 2.9905864393843022, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.731608390808105, "learning_rate": 1e-06, "loss": 0.4083, "mean_token_accuracy": 0.8716752529144287, "num_tokens": 896835113.0, "step": 23509 }, { "epoch": 2.9907136496628928, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.78073501586914, "learning_rate": 1e-06, "loss": 0.3901, "mean_token_accuracy": 0.8749051094055176, "num_tokens": 896866496.0, "step": 23510 }, { "epoch": 2.9908408599414833, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.687793731689453, "learning_rate": 1e-06, "loss": 0.4074, "mean_token_accuracy": 0.8734766244888306, "num_tokens": 896907535.0, "step": 23511 }, { "epoch": 2.990968070220074, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.746333122253418, "learning_rate": 1e-06, "loss": 0.3422, "mean_token_accuracy": 0.8920978903770447, "num_tokens": 896948574.0, "step": 23512 }, { "epoch": 2.9910952804986644, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.706768035888672, "learning_rate": 1e-06, "loss": 0.4171, "mean_token_accuracy": 0.870335578918457, "num_tokens": 896992377.0, "step": 23513 }, { "epoch": 2.991222490777255, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.78836727142334, "learning_rate": 1e-06, "loss": 0.4133, "mean_token_accuracy": 0.8697357177734375, "num_tokens": 897035411.0, "step": 23514 }, { "epoch": 2.9913497010558454, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.757246017456055, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.876958429813385, "num_tokens": 897074661.0, "step": 23515 }, { "epoch": 2.991476911334436, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.717061042785645, "learning_rate": 1e-06, "loss": 0.3988, "mean_token_accuracy": 0.8772299289703369, "num_tokens": 897115550.0, "step": 23516 }, { "epoch": 2.9916041216130265, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.797741889953613, "learning_rate": 1e-06, "loss": 0.4159, "mean_token_accuracy": 0.8689770102500916, "num_tokens": 897155091.0, "step": 23517 }, { "epoch": 2.991731331891617, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.66783618927002, "learning_rate": 1e-06, "loss": 0.4104, "mean_token_accuracy": 0.8696710467338562, "num_tokens": 897196551.0, "step": 23518 }, { "epoch": 2.991858542170207, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.723901748657227, "learning_rate": 1e-06, "loss": 0.359, "mean_token_accuracy": 0.8851696252822876, "num_tokens": 897231885.0, "step": 23519 }, { "epoch": 2.991985752448798, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.752398490905762, "learning_rate": 1e-06, "loss": 0.4005, "mean_token_accuracy": 0.875111997127533, "num_tokens": 897267944.0, "step": 23520 }, { "epoch": 2.992112962727388, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.70233154296875, "learning_rate": 1e-06, "loss": 0.3915, "mean_token_accuracy": 0.8775995969772339, "num_tokens": 897307330.0, "step": 23521 }, { "epoch": 2.992240173005979, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.725871086120605, "learning_rate": 1e-06, "loss": 0.3905, "mean_token_accuracy": 0.8772001266479492, "num_tokens": 897344028.0, "step": 23522 }, { "epoch": 2.992367383284569, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.761130332946777, "learning_rate": 1e-06, "loss": 0.4222, "mean_token_accuracy": 0.8671031594276428, "num_tokens": 897391213.0, "step": 23523 }, { "epoch": 2.99249459356316, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.726948738098145, "learning_rate": 1e-06, "loss": 0.4203, "mean_token_accuracy": 0.8686305284500122, "num_tokens": 897428222.0, "step": 23524 }, { "epoch": 2.9926218038417502, "ewc_loss": 0.040283203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 14.758790016174316, "learning_rate": 1e-06, "loss": 0.443, "mean_token_accuracy": 0.8646695613861084, "num_tokens": 897457737.0, "step": 23525 }, { "epoch": 2.992749014120341, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.718485832214355, "learning_rate": 1e-06, "loss": 0.3736, "mean_token_accuracy": 0.8817224502563477, "num_tokens": 897498925.0, "step": 23526 }, { "epoch": 2.9928762243989313, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.745563507080078, "learning_rate": 1e-06, "loss": 0.4152, "mean_token_accuracy": 0.8697198629379272, "num_tokens": 897533157.0, "step": 23527 }, { "epoch": 2.993003434677522, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.729856491088867, "learning_rate": 1e-06, "loss": 0.3909, "mean_token_accuracy": 0.8768036365509033, "num_tokens": 897574159.0, "step": 23528 }, { "epoch": 2.9931306449561124, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.755800247192383, "learning_rate": 1e-06, "loss": 0.4147, "mean_token_accuracy": 0.8666790723800659, "num_tokens": 897606970.0, "step": 23529 }, { "epoch": 2.993257855234703, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.715681076049805, "learning_rate": 1e-06, "loss": 0.4708, "mean_token_accuracy": 0.8526079654693604, "num_tokens": 897647725.0, "step": 23530 }, { "epoch": 2.9933850655132934, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.72541618347168, "learning_rate": 1e-06, "loss": 0.3867, "mean_token_accuracy": 0.8787071704864502, "num_tokens": 897687109.0, "step": 23531 }, { "epoch": 2.993512275791884, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.758066177368164, "learning_rate": 1e-06, "loss": 0.3695, "mean_token_accuracy": 0.8855384588241577, "num_tokens": 897720994.0, "step": 23532 }, { "epoch": 2.9936394860704745, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 14.742959976196289, "learning_rate": 1e-06, "loss": 0.3477, "mean_token_accuracy": 0.8913208246231079, "num_tokens": 897749855.0, "step": 23533 }, { "epoch": 2.993766696349065, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.666899681091309, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.8743751049041748, "num_tokens": 897792846.0, "step": 23534 }, { "epoch": 2.9938939066276555, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.737086296081543, "learning_rate": 1e-06, "loss": 0.3879, "mean_token_accuracy": 0.8776466846466064, "num_tokens": 897836765.0, "step": 23535 }, { "epoch": 2.994021116906246, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.659515380859375, "learning_rate": 1e-06, "loss": 0.3967, "mean_token_accuracy": 0.8752208948135376, "num_tokens": 897880918.0, "step": 23536 }, { "epoch": 2.9941483271848366, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.784454345703125, "learning_rate": 1e-06, "loss": 0.4, "mean_token_accuracy": 0.8733857870101929, "num_tokens": 897916676.0, "step": 23537 }, { "epoch": 2.994275537463427, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.65038013458252, "learning_rate": 1e-06, "loss": 0.3792, "mean_token_accuracy": 0.8805512189865112, "num_tokens": 897953968.0, "step": 23538 }, { "epoch": 2.9944027477420176, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.746098518371582, "learning_rate": 1e-06, "loss": 0.4125, "mean_token_accuracy": 0.8748996257781982, "num_tokens": 897993055.0, "step": 23539 }, { "epoch": 2.994529958020608, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.700910568237305, "learning_rate": 1e-06, "loss": 0.3694, "mean_token_accuracy": 0.8830415606498718, "num_tokens": 898030202.0, "step": 23540 }, { "epoch": 2.9946571682991987, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.724602699279785, "learning_rate": 1e-06, "loss": 0.4022, "mean_token_accuracy": 0.8706234693527222, "num_tokens": 898067234.0, "step": 23541 }, { "epoch": 2.9947843785777892, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.739479064941406, "learning_rate": 1e-06, "loss": 0.3797, "mean_token_accuracy": 0.8792283535003662, "num_tokens": 898105423.0, "step": 23542 }, { "epoch": 2.9949115888563798, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.707876205444336, "learning_rate": 1e-06, "loss": 0.3342, "mean_token_accuracy": 0.8938559889793396, "num_tokens": 898144550.0, "step": 23543 }, { "epoch": 2.99503879913497, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.73808765411377, "learning_rate": 1e-06, "loss": 0.3908, "mean_token_accuracy": 0.877830982208252, "num_tokens": 898179277.0, "step": 23544 }, { "epoch": 2.995166009413561, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.687458038330078, "learning_rate": 1e-06, "loss": 0.3922, "mean_token_accuracy": 0.8733447790145874, "num_tokens": 898218518.0, "step": 23545 }, { "epoch": 2.995293219692151, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.716948509216309, "learning_rate": 1e-06, "loss": 0.4299, "mean_token_accuracy": 0.8637810945510864, "num_tokens": 898259266.0, "step": 23546 }, { "epoch": 2.995420429970742, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.757866859436035, "learning_rate": 1e-06, "loss": 0.4278, "mean_token_accuracy": 0.8741587400436401, "num_tokens": 898292648.0, "step": 23547 }, { "epoch": 2.995547640249332, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.669085502624512, "learning_rate": 1e-06, "loss": 0.3953, "mean_token_accuracy": 0.879160463809967, "num_tokens": 898328413.0, "step": 23548 }, { "epoch": 2.995674850527923, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.692400932312012, "learning_rate": 1e-06, "loss": 0.397, "mean_token_accuracy": 0.8755654692649841, "num_tokens": 898376963.0, "step": 23549 }, { "epoch": 2.995802060806513, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.759003639221191, "learning_rate": 1e-06, "loss": 0.3584, "mean_token_accuracy": 0.8877561092376709, "num_tokens": 898407256.0, "step": 23550 }, { "epoch": 2.9959292710851035, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.705118179321289, "learning_rate": 1e-06, "loss": 0.3989, "mean_token_accuracy": 0.8778861165046692, "num_tokens": 898448800.0, "step": 23551 }, { "epoch": 2.996056481363694, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.656078338623047, "learning_rate": 1e-06, "loss": 0.4124, "mean_token_accuracy": 0.8719160556793213, "num_tokens": 898488912.0, "step": 23552 }, { "epoch": 2.9961836916422846, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.841900825500488, "learning_rate": 1e-06, "loss": 0.4176, "mean_token_accuracy": 0.868884801864624, "num_tokens": 898530434.0, "step": 23553 }, { "epoch": 2.996310901920875, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.683839797973633, "learning_rate": 1e-06, "loss": 0.3918, "mean_token_accuracy": 0.8762789964675903, "num_tokens": 898565044.0, "step": 23554 }, { "epoch": 2.9964381121994657, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.754109382629395, "learning_rate": 1e-06, "loss": 0.3689, "mean_token_accuracy": 0.8854598999023438, "num_tokens": 898601798.0, "step": 23555 }, { "epoch": 2.996565322478056, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.612119674682617, "learning_rate": 1e-06, "loss": 0.4254, "mean_token_accuracy": 0.8659152984619141, "num_tokens": 898644789.0, "step": 23556 }, { "epoch": 2.9966925327566467, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.742822647094727, "learning_rate": 1e-06, "loss": 0.3978, "mean_token_accuracy": 0.8752719163894653, "num_tokens": 898683928.0, "step": 23557 }, { "epoch": 2.9968197430352372, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.714101791381836, "learning_rate": 1e-06, "loss": 0.4097, "mean_token_accuracy": 0.8704062104225159, "num_tokens": 898726217.0, "step": 23558 }, { "epoch": 2.9969469533138278, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.779632568359375, "learning_rate": 1e-06, "loss": 0.3664, "mean_token_accuracy": 0.8850125074386597, "num_tokens": 898760575.0, "step": 23559 }, { "epoch": 2.9970741635924183, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.716609001159668, "learning_rate": 1e-06, "loss": 0.4043, "mean_token_accuracy": 0.8716838359832764, "num_tokens": 898801232.0, "step": 23560 }, { "epoch": 2.997201373871009, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.77733325958252, "learning_rate": 1e-06, "loss": 0.3726, "mean_token_accuracy": 0.8813820481300354, "num_tokens": 898837174.0, "step": 23561 }, { "epoch": 2.9973285841495994, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.781933784484863, "learning_rate": 1e-06, "loss": 0.4325, "mean_token_accuracy": 0.8673704266548157, "num_tokens": 898875849.0, "step": 23562 }, { "epoch": 2.99745579442819, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 14.671784400939941, "learning_rate": 1e-06, "loss": 0.3739, "mean_token_accuracy": 0.8837389945983887, "num_tokens": 898909975.0, "step": 23563 }, { "epoch": 2.9975830047067804, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.764569282531738, "learning_rate": 1e-06, "loss": 0.416, "mean_token_accuracy": 0.8687890768051147, "num_tokens": 898948250.0, "step": 23564 }, { "epoch": 2.997710214985371, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.747814178466797, "learning_rate": 1e-06, "loss": 0.3605, "mean_token_accuracy": 0.8917535543441772, "num_tokens": 898989122.0, "step": 23565 }, { "epoch": 2.9978374252639615, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.795440673828125, "learning_rate": 1e-06, "loss": 0.4089, "mean_token_accuracy": 0.8705965280532837, "num_tokens": 899028715.0, "step": 23566 }, { "epoch": 2.997964635542552, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.711880683898926, "learning_rate": 1e-06, "loss": 0.3835, "mean_token_accuracy": 0.8780949711799622, "num_tokens": 899064101.0, "step": 23567 }, { "epoch": 2.9980918458211425, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.743969917297363, "learning_rate": 1e-06, "loss": 0.3917, "mean_token_accuracy": 0.8788621425628662, "num_tokens": 899101759.0, "step": 23568 }, { "epoch": 2.9982190560997326, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.718925476074219, "learning_rate": 1e-06, "loss": 0.4148, "mean_token_accuracy": 0.8691400289535522, "num_tokens": 899142273.0, "step": 23569 }, { "epoch": 2.9983462663783236, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.763026237487793, "learning_rate": 1e-06, "loss": 0.3923, "mean_token_accuracy": 0.8772043585777283, "num_tokens": 899175878.0, "step": 23570 }, { "epoch": 2.9984734766569137, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.692209243774414, "learning_rate": 1e-06, "loss": 0.3563, "mean_token_accuracy": 0.8861227631568909, "num_tokens": 899208937.0, "step": 23571 }, { "epoch": 2.9986006869355046, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.765946388244629, "learning_rate": 1e-06, "loss": 0.3979, "mean_token_accuracy": 0.874297022819519, "num_tokens": 899244088.0, "step": 23572 }, { "epoch": 2.9987278972140947, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.698348045349121, "learning_rate": 1e-06, "loss": 0.4189, "mean_token_accuracy": 0.8709908127784729, "num_tokens": 899280111.0, "step": 23573 }, { "epoch": 2.9988551074926852, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.782179832458496, "learning_rate": 1e-06, "loss": 0.3997, "mean_token_accuracy": 0.8754109144210815, "num_tokens": 899310882.0, "step": 23574 }, { "epoch": 2.9989823177712758, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.694751739501953, "learning_rate": 1e-06, "loss": 0.4502, "mean_token_accuracy": 0.8625215291976929, "num_tokens": 899346820.0, "step": 23575 }, { "epoch": 2.9991095280498663, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.75478744506836, "learning_rate": 1e-06, "loss": 0.3561, "mean_token_accuracy": 0.8887381553649902, "num_tokens": 899382044.0, "step": 23576 }, { "epoch": 2.999236738328457, "ewc_loss": 0.040771484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 14.681840896606445, "learning_rate": 1e-06, "loss": 0.3876, "mean_token_accuracy": 0.8762622475624084, "num_tokens": 899418389.0, "step": 23577 }, { "epoch": 2.9993639486070474, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.733728408813477, "learning_rate": 1e-06, "loss": 0.411, "mean_token_accuracy": 0.8734946250915527, "num_tokens": 899458486.0, "step": 23578 }, { "epoch": 2.999491158885638, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.682801246643066, "learning_rate": 1e-06, "loss": 0.4037, "mean_token_accuracy": 0.8740428686141968, "num_tokens": 899498168.0, "step": 23579 }, { "epoch": 2.9996183691642284, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 14.72769832611084, "learning_rate": 1e-06, "loss": 0.3472, "mean_token_accuracy": 0.8921084403991699, "num_tokens": 899538655.0, "step": 23580 }, { "epoch": 2.999745579442819, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 14.767566680908203, "learning_rate": 1e-06, "loss": 0.4289, "mean_token_accuracy": 0.8640438914299011, "num_tokens": 899581693.0, "step": 23581 }, { "epoch": 2.9998727897214095, "ewc_loss": 0.04052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 14.714250564575195, "learning_rate": 1e-06, "loss": 0.4076, "mean_token_accuracy": 0.8724070191383362, "num_tokens": 899623817.0, "step": 23582 }, { "epoch": 3.0, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 14.696348190307617, "learning_rate": 1e-06, "loss": 0.387, "mean_token_accuracy": 0.8792088031768799, "num_tokens": 899664226.0, "step": 23583 }, { "epoch": 3.0, "ewc_loss": 0.041015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 2.682209014892578e-05, "step": 23583, "total_flos": 5.62815163329864e+19, "train_loss": 0.42514991610376646, "train_runtime": 47075.3517, "train_samples_per_second": 8.015, "train_steps_per_second": 0.501 } ], "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 }